core.py

import binascii
import hashlib
import itertools
import math
import operator
import re
from functools import reduce
from typing import Any

import numpy as np
from numcodecs.compat import ensure_bytes

from zarr._storage.store import _prefix_to_attrs_key, assert_zarr_v3_api_available
from zarr.attrs import Attributes
from zarr.codecs import AsType, get_codec
from zarr.context import Context
from zarr.errors import ArrayNotFoundError, ReadOnlyError, ArrayIndexError
from zarr.indexing import (
    BasicIndexer,
    CoordinateIndexer,
    MaskIndexer,
    OIndex,
    OrthogonalIndexer,
    VIndex,
    PartialChunkIterator,
    check_fields,
    check_no_multi_fields,
    ensure_tuple,
    err_too_many_indices,
    is_contiguous_selection,
    is_pure_fancy_indexing,
    is_pure_orthogonal_indexing,
    is_scalar,
    pop_fields,
)
from zarr.storage import (
    _get_hierarchy_metadata,
    _prefix_to_array_key,
    KVStore,
    getsize,
    listdir,
    normalize_store_arg,
)
from zarr.util import (
    ConstantMap,
    all_equal,
    InfoReporter,
    check_array_shape,
    human_readable_size,
    is_total_slice,
    nolock,
    normalize_chunks,
    normalize_resize_args,
    normalize_shape,
    normalize_storage_path,
    PartialReadBuffer,
    UncompressedPartialReadBufferV3,
    ensure_ndarray_like,
)


# noinspection PyUnresolvedReferences
class Array:
    """Instantiate an array from an initialized store.

    Parameters
    ----------
    store : MutableMapping
        Array store, already initialized.
    path : string, optional
        Storage path.
    read_only : bool, optional
        True if array should be protected against modification.
    chunk_store : MutableMapping, optional
        Separate storage for chunks. If not provided, `store` will be used
        for storage of both chunks and metadata.
    synchronizer : object, optional
        Array synchronizer.
    cache_metadata : bool, optional
        If True (default), array configuration metadata will be cached for the
        lifetime of the object. If False, array metadata will be reloaded
        prior to all data access and modification operations (may incur
        overhead depending on storage and data access pattern).
    cache_attrs : bool, optional
        If True (default), user attributes will be cached for attribute read
        operations. If False, user attributes are reloaded from the store prior
        to all attribute read operations.
    partial_decompress : bool, optional
        If True and while the chunk_store is a FSStore and the compression used
        is Blosc, when getting data from the array chunks will be partially
        read and decompressed when possible.

        .. versionadded:: 2.7

    write_empty_chunks : bool, optional
        If True, all chunks will be stored regardless of their contents. If
        False (default), each chunk is compared to the array's fill value prior
        to storing. If a chunk is uniformly equal to the fill value, then that
        chunk is not be stored, and the store entry for that chunk's key is
        deleted. This setting enables sparser storage, as only chunks with
        non-fill-value data are stored, at the expense of overhead associated
        with checking the data of each chunk.

        .. versionadded:: 2.11

    meta_array : array-like, optional
        An array instance to use for determining arrays to create and return
        to users. Use `numpy.empty(())` by default.

        .. versionadded:: 2.13


    Attributes
    ----------
    store
    path
    name
    read_only
    chunk_store
    shape
    chunks
    dtype
    compression
    compression_opts
    dimension_separator
    fill_value
    order
    synchronizer
    filters
    attrs
    size
    itemsize
    nbytes
    nbytes_stored
    cdata_shape
    nchunks
    nchunks_initialized
    is_view
    info
    vindex
    oindex
    write_empty_chunks
    meta_array

    Methods
    -------
    __getitem__
    __setitem__
    get_basic_selection
    set_basic_selection
    get_orthogonal_selection
    set_orthogonal_selection
    get_mask_selection
    set_mask_selection
    get_coordinate_selection
    set_coordinate_selection
    digest
    hexdigest
    resize
    append
    view
    astype

    """

    def __init__(
        self,
        store: Any,  # BaseStore not strictly required due to normalize_store_arg
        path=None,
        read_only=False,
        chunk_store=None,
        synchronizer=None,
        cache_metadata=True,
        cache_attrs=True,
        partial_decompress=False,
        write_empty_chunks=True,
        zarr_version=None,
        meta_array=None,
    ):
        # N.B., expect at this point store is fully initialized with all
        # configuration metadata fully specified and normalized

        store = normalize_store_arg(store, zarr_version=zarr_version)
        if zarr_version is None:
            zarr_version = store._store_version

        if zarr_version != 2:
            assert_zarr_v3_api_available()

        if chunk_store is not None:
            chunk_store = normalize_store_arg(chunk_store,
                                              zarr_version=zarr_version)

        self._store = store
        self._chunk_store = chunk_store
        self._transformed_chunk_store = None
        self._path = normalize_storage_path(path)
        if self._path:
            self._key_prefix = self._path + '/'
        else:
            self._key_prefix = ''
        self._read_only = bool(read_only)
        self._synchronizer = synchronizer
        self._cache_metadata = cache_metadata
        self._is_view = False
        self._partial_decompress = partial_decompress
        self._write_empty_chunks = write_empty_chunks
        if meta_array is not None:
            self._meta_array = np.empty_like(meta_array, shape=())
        else:
            self._meta_array = np.empty(())
        self._version = zarr_version
        if self._version == 3:
            self._data_key_prefix = 'data/root/' + self._key_prefix
            self._data_path = 'data/root/' + self._path
            self._hierarchy_metadata = _get_hierarchy_metadata(store=self._store)
            self._metadata_key_suffix = self._hierarchy_metadata['metadata_key_suffix']

        # initialize metadata
        self._load_metadata()

        # initialize attributes
        akey = _prefix_to_attrs_key(self._store, self._key_prefix)
        self._attrs = Attributes(store, key=akey, read_only=read_only,
                                 synchronizer=synchronizer, cache=cache_attrs)

        # initialize info reporter
        self._info_reporter = InfoReporter(self)

        # initialize indexing helpers
        self._oindex = OIndex(self)
        self._vindex = VIndex(self)

    def _load_metadata(self):
        """(Re)load metadata from store."""
        if self._synchronizer is None:
            self._load_metadata_nosync()
        else:
            mkey = _prefix_to_array_key(self._store, self._key_prefix)
            with self._synchronizer[mkey]:
                self._load_metadata_nosync()

    def _load_metadata_nosync(self):
        try:
            mkey = _prefix_to_array_key(self._store, self._key_prefix)
            meta_bytes = self._store[mkey]
        except KeyError:
            raise ArrayNotFoundError(self._path)
        else:

            # decode and store metadata as instance members
            meta = self._store._metadata_class.decode_array_metadata(meta_bytes)
            self._meta = meta
            self._shape = meta['shape']
            self._fill_value = meta['fill_value']
            dimension_separator = meta.get('dimension_separator', None)
            if self._version == 2:
                self._chunks = meta['chunks']
                self._dtype = meta['dtype']
                self._order = meta['order']
                if dimension_separator is None:
                    try:
                        dimension_separator = self._store._dimension_separator
                    except (AttributeError, KeyError):
                        pass

                    # Fallback for any stores which do not choose a default
                    if dimension_separator is None:
                        dimension_separator = "."
            else:
                self._chunks = meta['chunk_grid']['chunk_shape']
                self._dtype = meta['data_type']
                self._order = meta['chunk_memory_layout']
                chunk_separator = meta['chunk_grid']['separator']
                if dimension_separator is None:
                    dimension_separator = meta.get('dimension_separator', chunk_separator)

            self._dimension_separator = dimension_separator

            # setup compressor
            compressor = meta.get('compressor', None)
            if compressor is None:
                self._compressor = None
            elif self._version == 2:
                self._compressor = get_codec(compressor)
            else:
                self._compressor = compressor

            # setup filters
            if self._version == 2:
                filters = meta.get('filters', [])
            else:
                # TODO: storing filters under attributes for now since the v3
                #       array metadata does not have a 'filters' attribute.
                filters = meta['attributes'].get('filters', [])
            if filters:
                filters = [get_codec(config) for config in filters]
            self._filters = filters

            if self._version == 3:
                storage_transformers = meta.get('storage_transformers', [])
                if storage_transformers:
                    transformed_store = self._chunk_store or self._store
                    for storage_transformer in storage_transformers[::-1]:
                        transformed_store = storage_transformer._copy_for_array(
                            self, transformed_store
                        )
                    self._transformed_chunk_store = transformed_store

    def _refresh_metadata(self):
        if not self._cache_metadata:
            self._load_metadata()

    def _refresh_metadata_nosync(self):
        if not self._cache_metadata and not self._is_view:
            self._load_metadata_nosync()

    def _flush_metadata_nosync(self):
        if self._is_view:
            raise PermissionError('operation not permitted for views')

        if self._compressor:
            compressor_config = self._compressor.get_config()
        else:
            compressor_config = None
        if self._filters:
            filters_config = [f.get_config() for f in self._filters]
        else:
            filters_config = None
        _compressor = compressor_config if self._version == 2 else self._compressor
        meta = dict(shape=self._shape, compressor=_compressor,
                    fill_value=self._fill_value, filters=filters_config)
        if getattr(self._store, '_store_version', 2) == 2:
            meta.update(
                dict(chunks=self._chunks, dtype=self._dtype, order=self._order)
            )
        else:
            meta.update(
                dict(chunk_grid=dict(type='regular',
                                     chunk_shape=self._chunks,
                                     separator=self._dimension_separator),
                     data_type=self._dtype,
                     chunk_memory_layout=self._order,
                     attributes=self.attrs.asdict())
            )
        mkey = _prefix_to_array_key(self._store, self._key_prefix)
        self._store[mkey] = self._store._metadata_class.encode_array_metadata(meta)

    @property
    def store(self):
        """A MutableMapping providing the underlying storage for the array."""
        return self._store

    @property
    def path(self):
        """Storage path."""
        return self._path

    @property
    def name(self):
        """Array name following h5py convention."""
        if self.path:
            # follow h5py convention: add leading slash
            name = self.path
            if name[0] != '/':
                name = '/' + name
            return name
        return None

    @property
    def basename(self):
        """Final component of name."""
        if self.name is not None:
            return self.name.split('/')[-1]
        return None

    @property
    def read_only(self):
        """A boolean, True if modification operations are not permitted."""
        return self._read_only

    @read_only.setter
    def read_only(self, value):
        self._read_only = bool(value)

    @property
    def chunk_store(self):
        """A MutableMapping providing the underlying storage for array chunks."""
        if self._transformed_chunk_store is not None:
            return self._transformed_chunk_store
        elif self._chunk_store is not None:
            return self._chunk_store
        else:
            return self._store

    @property
    def shape(self):
        """A tuple of integers describing the length of each dimension of
        the array."""
        # N.B., shape may change if array is resized, hence need to refresh
        # metadata
        self._refresh_metadata()
        return self._shape

    @shape.setter
    def shape(self, value):
        self.resize(value)

    @property
    def chunks(self):
        """A tuple of integers describing the length of each dimension of a
        chunk of the array."""
        return self._chunks

    @property
    def dtype(self):
        """The NumPy data type."""
        return self._dtype

    @property
    def compressor(self):
        """Primary compression codec."""
        return self._compressor

    @property
    def fill_value(self):
        """A value used for uninitialized portions of the array."""
        return self._fill_value

    @fill_value.setter
    def fill_value(self, new):
        self._fill_value = new
        self._flush_metadata_nosync()

    @property
    def order(self):
        """A string indicating the order in which bytes are arranged within
        chunks of the array."""
        return self._order

    @property
    def filters(self):
        """One or more codecs used to transform data prior to compression."""
        return self._filters

    @property
    def synchronizer(self):
        """Object used to synchronize write access to the array."""
        return self._synchronizer

    @property
    def attrs(self):
        """A MutableMapping containing user-defined attributes. Note that
        attribute values must be JSON serializable."""
        return self._attrs

    @property
    def ndim(self):
        """Number of dimensions."""
        return len(self._shape)

    @property
    def _size(self):
        return reduce(operator.mul, self._shape, 1)

    @property
    def size(self):
        """The total number of elements in the array."""
        # N.B., this property depends on shape, and shape may change if array
        # is resized, hence need to refresh metadata
        self._refresh_metadata()
        return self._size

    @property
    def itemsize(self):
        """The size in bytes of each item in the array."""
        return self.dtype.itemsize

    @property
    def _nbytes(self):
        return self._size * self.itemsize

    @property
    def nbytes(self):
        """The total number of bytes that would be required to store the
        array without compression."""
        # N.B., this property depends on shape, and shape may change if array
        # is resized, hence need to refresh metadata
        self._refresh_metadata()
        return self._nbytes

    @property
    def nbytes_stored(self):
        """The total number of stored bytes of data for the array. This
        includes storage required for configuration metadata and user
        attributes."""
        m = getsize(self._store, self._path)
        if self._chunk_store is None:
            return m
        else:
            n = getsize(self._chunk_store, self._path)
            if m < 0 or n < 0:
                return -1
            else:
                return m + n

    @property
    def _cdata_shape(self):
        if self._shape == ():
            return 1,
        else:
            return tuple(math.ceil(s / c)
                         for s, c in zip(self._shape, self._chunks))

    @property
    def cdata_shape(self):
        """A tuple of integers describing the number of chunks along each
        dimension of the array."""
        self._refresh_metadata()
        return self._cdata_shape

    @property
    def _nchunks(self):
        return reduce(operator.mul, self._cdata_shape, 1)

    @property
    def nchunks(self):
        """Total number of chunks."""
        self._refresh_metadata()
        return self._nchunks

    @property
    def nchunks_initialized(self):
        """The number of chunks that have been initialized with some data."""

        # count chunk keys
        if self._version == 3:
            # # key pattern for chunk keys
            # prog = re.compile(r'\.'.join([r'c\d+'] * min(1, self.ndim)))
            # # get chunk keys, excluding the prefix
            # members = self.chunk_store.list_prefix(self._data_path)
            # members = [k.split(self._data_key_prefix)[1] for k in members]
            # # count the chunk keys
            # return sum(1 for k in members if prog.match(k))

            # key pattern for chunk keys
            prog = re.compile(self._data_key_prefix + r'c\d+')  # TODO: ndim == 0 case?
            # get chunk keys, excluding the prefix
            members = self.chunk_store.list_prefix(self._data_path)
            # count the chunk keys
            return sum(1 for k in members if prog.match(k))
        else:
            # key pattern for chunk keys
            prog = re.compile(r'\.'.join([r'\d+'] * min(1, self.ndim)))

            # count chunk keys
            return sum(1 for k in listdir(self.chunk_store, self._path) if prog.match(k))

    # backwards compatibility
    initialized = nchunks_initialized

    @property
    def is_view(self):
        """A boolean, True if this array is a view on another array."""
        return self._is_view

    @property
    def oindex(self):
        """Shortcut for orthogonal (outer) indexing, see :func:`get_orthogonal_selection` and
        :func:`set_orthogonal_selection` for documentation and examples."""
        return self._oindex

    @property
    def vindex(self):
        """Shortcut for vectorized (inner) indexing, see :func:`get_coordinate_selection`,
        :func:`set_coordinate_selection`, :func:`get_mask_selection` and
        :func:`set_mask_selection` for documentation and examples."""
        return self._vindex

    @property
    def write_empty_chunks(self) -> bool:
        """A Boolean, True if chunks composed of the array's fill value
        will be stored. If False, such chunks will not be stored.
        """
        return self._write_empty_chunks

    @property
    def meta_array(self):
        """An array-like instance to use for determining arrays to create and return
        to users.
        """
        return self._meta_array

    def __eq__(self, other):
        return (
            isinstance(other, Array) and
            self.store == other.store and
            self.read_only == other.read_only and
            self.path == other.path and
            not self._is_view
            # N.B., no need to compare other properties, should be covered by
            # store comparison
        )

    def __array__(self, *args):
        a = self[...]
        if args:
            a = a.astype(args[0])
        return a

    def islice(self, start=None, end=None):
        """
        Yield a generator for iterating over the entire or parts of the
        array. Uses a cache so chunks only have to be decompressed once.

        Parameters
        ----------
        start : int, optional
            Start index for the generator to start at. Defaults to 0.
        end : int, optional
            End index for the generator to stop at. Defaults to self.shape[0].

        Yields
        ------
        out : generator
            A generator that can be used to iterate over the requested region
            the array.

        Examples
        --------
        Setup a 1-dimensional array::

            >>> import zarr
            >>> import numpy as np
            >>> z = zarr.array(np.arange(100))

        Iterate over part of the array:
            >>> for value in z.islice(25, 30): value;
            25
            26
            27
            28
            29
        """

        if len(self.shape) == 0:
            # Same error as numpy
            raise TypeError("iteration over a 0-d array")
        if start is None:
            start = 0
        if end is None or end > self.shape[0]:
            end = self.shape[0]

        if not isinstance(start, int) or start < 0:
            raise ValueError('start must be a nonnegative integer')

        if not isinstance(end, int) or end < 0:
            raise ValueError('end must be a nonnegative integer')

        # Avoid repeatedly decompressing chunks by iterating over the chunks
        # in the first dimension.
        chunk_size = self.chunks[0]
        chunk = None
        for j in range(start, end):
            if j % chunk_size == 0:
                chunk = self[j: j + chunk_size]
            # init chunk if we start offset of chunk borders
            elif chunk is None:
                chunk_start = j - j % chunk_size
                chunk_end = chunk_start + chunk_size
                chunk = self[chunk_start:chunk_end]
            yield chunk[j % chunk_size]

    def __iter__(self):
        return self.islice()

    def __len__(self):
        if self.shape:
            return self.shape[0]
        else:
            # 0-dimensional array, same error message as numpy
            raise TypeError('len() of unsized object')

    def __getitem__(self, selection):
        """Retrieve data for an item or region of the array.

        Parameters
        ----------
        selection : tuple
            An integer index or slice or tuple of int/slice objects specifying the
            requested item or region for each dimension of the array.

        Returns
        -------
        out : ndarray
            A NumPy array containing the data for the requested region.

        Examples
        --------
        Setup a 1-dimensional array::

            >>> import zarr
            >>> import numpy as np
            >>> z = zarr.array(np.arange(100))

        Retrieve a single item::

            >>> z[5]
            5

        Retrieve a region via slicing::

            >>> z[:5]
            array([0, 1, 2, 3, 4])
            >>> z[-5:]
            array([95, 96, 97, 98, 99])
            >>> z[5:10]
            array([5, 6, 7, 8, 9])
            >>> z[5:10:2]
            array([5, 7, 9])
            >>> z[::2]
            array([ 0,  2,  4, ..., 94, 96, 98])

        Load the entire array into memory::

            >>> z[...]
            array([ 0,  1,  2, ..., 97, 98, 99])

        Setup a 2-dimensional array::

            >>> z = zarr.array(np.arange(100).reshape(10, 10))

        Retrieve an item::

            >>> z[2, 2]
            22

        Retrieve a region via slicing::

            >>> z[1:3, 1:3]
            array([[11, 12],
                   [21, 22]])
            >>> z[1:3, :]
            array([[10, 11, 12, 13, 14, 15, 16, 17, 18, 19],
                   [20, 21, 22, 23, 24, 25, 26, 27, 28, 29]])
            >>> z[:, 1:3]
            array([[ 1,  2],
                   [11, 12],
                   [21, 22],
                   [31, 32],
                   [41, 42],
                   [51, 52],
                   [61, 62],
                   [71, 72],
                   [81, 82],
                   [91, 92]])
            >>> z[0:5:2, 0:5:2]
            array([[ 0,  2,  4],
                   [20, 22, 24],
                   [40, 42, 44]])
            >>> z[::2, ::2]
            array([[ 0,  2,  4,  6,  8],
                   [20, 22, 24, 26, 28],
                   [40, 42, 44, 46, 48],
                   [60, 62, 64, 66, 68],
                   [80, 82, 84, 86, 88]])

        Load the entire array into memory::

            >>> z[...]
            array([[ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9],
                   [10, 11, 12, 13, 14, 15, 16, 17, 18, 19],
                   [20, 21, 22, 23, 24, 25, 26, 27, 28, 29],
                   [30, 31, 32, 33, 34, 35, 36, 37, 38, 39],
                   [40, 41, 42, 43, 44, 45, 46, 47, 48, 49],
                   [50, 51, 52, 53, 54, 55, 56, 57, 58, 59],
                   [60, 61, 62, 63, 64, 65, 66, 67, 68, 69],
                   [70, 71, 72, 73, 74, 75, 76, 77, 78, 79],
                   [80, 81, 82, 83, 84, 85, 86, 87, 88, 89],
                   [90, 91, 92, 93, 94, 95, 96, 97, 98, 99]])

        For arrays with a structured dtype, specific fields can be retrieved, e.g.::

            >>> a = np.array([(b'aaa', 1, 4.2),
            ...               (b'bbb', 2, 8.4),
            ...               (b'ccc', 3, 12.6)],
            ...              dtype=[('foo', 'S3'), ('bar', 'i4'), ('baz', 'f8')])
            >>> z = zarr.array(a)
            >>> z['foo']
            array([b'aaa', b'bbb', b'ccc'],
                  dtype='|S3')

        Notes
        -----
        Slices with step > 1 are supported, but slices with negative step are not.

        Currently the implementation for __getitem__ is provided by
        :func:`vindex` if the indexing is pure fancy indexing (ie a
        broadcast-compatible tuple of integer array indices), or by
        :func:`set_basic_selection` otherwise.

        Effectively, this means that the following indexing modes are supported:

           - integer indexing
           - slice indexing
           - mixed slice and integer indexing
           - boolean indexing
           - fancy indexing (vectorized list of integers)

        For specific indexing options including outer indexing, see the
        methods listed under See Also.

        See Also
        --------
        get_basic_selection, set_basic_selection, get_mask_selection, set_mask_selection,
        get_coordinate_selection, set_coordinate_selection, get_orthogonal_selection,
        set_orthogonal_selection, vindex, oindex, __setitem__

        """
        fields, pure_selection = pop_fields(selection)
        if is_pure_fancy_indexing(pure_selection, self.ndim):
            result = self.vindex[selection]
        elif is_pure_orthogonal_indexing(pure_selection, self.ndim):
            result = self.get_orthogonal_selection(pure_selection, fields=fields)
        else:
            result = self.get_basic_selection(pure_selection, fields=fields)
        return result

    def get_basic_selection(self, selection=Ellipsis, out=None, fields=None):
        """Retrieve data for an item or region of the array.

        Parameters
        ----------
        selection : tuple
            A tuple specifying the requested item or region for each dimension of the
            array. May be any combination of int and/or slice for multidimensional arrays.
        out : ndarray, optional
            If given, load the selected data directly into this array.
        fields : str or sequence of str, optional
            For arrays with a structured dtype, one or more fields can be specified to
            extract data for.

        Returns
        -------
        out : ndarray
            A NumPy array containing the data for the requested region.

        Examples
        --------
        Setup a 1-dimensional array::

            >>> import zarr
            >>> import numpy as np
            >>> z = zarr.array(np.arange(100))

        Retrieve a single item::

            >>> z.get_basic_selection(5)
            5

        Retrieve a region via slicing::

            >>> z.get_basic_selection(slice(5))
            array([0, 1, 2, 3, 4])
            >>> z.get_basic_selection(slice(-5, None))
            array([95, 96, 97, 98, 99])
            >>> z.get_basic_selection(slice(5, 10))
            array([5, 6, 7, 8, 9])
            >>> z.get_basic_selection(slice(5, 10, 2))
            array([5, 7, 9])
            >>> z.get_basic_selection(slice(None, None, 2))
            array([  0,  2,  4, ..., 94, 96, 98])

        Setup a 2-dimensional array::

            >>> z = zarr.array(np.arange(100).reshape(10, 10))

        Retrieve an item::

            >>> z.get_basic_selection((2, 2))
            22

        Retrieve a region via slicing::

            >>> z.get_basic_selection((slice(1, 3), slice(1, 3)))
            array([[11, 12],
                   [21, 22]])
            >>> z.get_basic_selection((slice(1, 3), slice(None)))
            array([[10, 11, 12, 13, 14, 15, 16, 17, 18, 19],
                   [20, 21, 22, 23, 24, 25, 26, 27, 28, 29]])
            >>> z.get_basic_selection((slice(None), slice(1, 3)))
            array([[ 1,  2],
                   [11, 12],
                   [21, 22],
                   [31, 32],
                   [41, 42],
                   [51, 52],
                   [61, 62],
                   [71, 72],
                   [81, 82],
                   [91, 92]])
            >>> z.get_basic_selection((slice(0, 5, 2), slice(0, 5, 2)))
            array([[ 0,  2,  4],
                   [20, 22, 24],
                   [40, 42, 44]])
            >>> z.get_basic_selection((slice(None, None, 2), slice(None, None, 2)))
            array([[ 0,  2,  4,  6,  8],
                   [20, 22, 24, 26, 28],
                   [40, 42, 44, 46, 48],
                   [60, 62, 64, 66, 68],
                   [80, 82, 84, 86, 88]])

        For arrays with a structured dtype, specific fields can be retrieved, e.g.::

            >>> a = np.array([(b'aaa', 1, 4.2),
            ...               (b'bbb', 2, 8.4),
            ...               (b'ccc', 3, 12.6)],
            ...              dtype=[('foo', 'S3'), ('bar', 'i4'), ('baz', 'f8')])
            >>> z = zarr.array(a)
            >>> z.get_basic_selection(slice(2), fields='foo')
            array([b'aaa', b'bbb'],
                  dtype='|S3')

        Notes
        -----
        Slices with step > 1 are supported, but slices with negative step are not.

        Currently this method provides the implementation for accessing data via the
        square bracket notation (__getitem__). See :func:`__getitem__` for examples
        using the alternative notation.

        See Also
        --------
        set_basic_selection, get_mask_selection, set_mask_selection,
        get_coordinate_selection, set_coordinate_selection, get_orthogonal_selection,
        set_orthogonal_selection, vindex, oindex, __getitem__, __setitem__

        """

        # refresh metadata
        if not self._cache_metadata:
            self._load_metadata()

        # check args
        check_fields(fields, self._dtype)

        # handle zero-dimensional arrays
        if self._shape == ():
            return self._get_basic_selection_zd(selection=selection, out=out,
                                                fields=fields)
        else:
            return self._get_basic_selection_nd(selection=selection, out=out,
                                                fields=fields)

    def _get_basic_selection_zd(self, selection, out=None, fields=None):
        # special case basic selection for zero-dimensional array

        # check selection is valid
        selection = ensure_tuple(selection)
        if selection not in ((), (Ellipsis,)):
            err_too_many_indices(selection, ())

        try:
            # obtain encoded data for chunk
            ckey = self._chunk_key((0,))
            cdata = self.chunk_store[ckey]

        except KeyError:
            # chunk not initialized
            chunk = np.zeros_like(self._meta_array, shape=(), dtype=self._dtype)
            if self._fill_value is not None:
                chunk.fill(self._fill_value)

        else:
            chunk = self._decode_chunk(cdata)

        # handle fields
        if fields:
            chunk = chunk[fields]

        # handle selection of the scalar value via empty tuple
        if out is None:
            out = chunk[selection]
        else:
            out[selection] = chunk[selection]

        return out

    def _get_basic_selection_nd(self, selection, out=None, fields=None):
        # implementation of basic selection for array with at least one dimension

        # setup indexer
        indexer = BasicIndexer(selection, self)

        return self._get_selection(indexer=indexer, out=out, fields=fields)

    def get_orthogonal_selection(self, selection, out=None, fields=None):
        """Retrieve data by making a selection for each dimension of the array. For
        example, if an array has 2 dimensions, allows selecting specific rows and/or
        columns. The selection for each dimension can be either an integer (indexing a
        single item), a slice, an array of integers, or a Boolean array where True
        values indicate a selection.

        Parameters
        ----------
        selection : tuple
            A selection for each dimension of the array. May be any combination of int,
            slice, integer array or Boolean array.
        out : ndarray, optional
            If given, load the selected data directly into this array.
        fields : str or sequence of str, optional
            For arrays with a structured dtype, one or more fields can be specified to
            extract data for.

        Returns
        -------
        out : ndarray
            A NumPy array containing the data for the requested selection.

        Examples
        --------
        Setup a 2-dimensional array::

            >>> import zarr
            >>> import numpy as np
            >>> z = zarr.array(np.arange(100).reshape(10, 10))

        Retrieve rows and columns via any combination of int, slice, integer array and/or
        Boolean array::

            >>> z.get_orthogonal_selection(([1, 4], slice(None)))
            array([[10, 11, 12, 13, 14, 15, 16, 17, 18, 19],
                   [40, 41, 42, 43, 44, 45, 46, 47, 48, 49]])
            >>> z.get_orthogonal_selection((slice(None), [1, 4]))
            array([[ 1,  4],
                   [11, 14],
                   [21, 24],
                   [31, 34],
                   [41, 44],
                   [51, 54],
                   [61, 64],
                   [71, 74],
                   [81, 84],
                   [91, 94]])
            >>> z.get_orthogonal_selection(([1, 4], [1, 4]))
            array([[11, 14],
                   [41, 44]])
            >>> sel = np.zeros(z.shape[0], dtype=bool)
            >>> sel[1] = True
            >>> sel[4] = True
            >>> z.get_orthogonal_selection((sel, sel))
            array([[11, 14],
                   [41, 44]])

        For convenience, the orthogonal selection functionality is also available via the
        `oindex` property, e.g.::

            >>> z.oindex[[1, 4], :]
            array([[10, 11, 12, 13, 14, 15, 16, 17, 18, 19],
                   [40, 41, 42, 43, 44, 45, 46, 47, 48, 49]])
            >>> z.oindex[:, [1, 4]]
            array([[ 1,  4],
                   [11, 14],
                   [21, 24],
                   [31, 34],
                   [41, 44],
                   [51, 54],
                   [61, 64],
                   [71, 74],
                   [81, 84],
                   [91, 94]])
            >>> z.oindex[[1, 4], [1, 4]]
            array([[11, 14],
                   [41, 44]])
            >>> sel = np.zeros(z.shape[0], dtype=bool)
            >>> sel[1] = True
            >>> sel[4] = True
            >>> z.oindex[sel, sel]
            array([[11, 14],
                   [41, 44]])

        Notes
        -----
        Orthogonal indexing is also known as outer indexing.

        Slices with step > 1 are supported, but slices with negative step are not.

        See Also
        --------
        get_basic_selection, set_basic_selection, get_mask_selection, set_mask_selection,
        get_coordinate_selection, set_coordinate_selection, set_orthogonal_selection,
        vindex, oindex, __getitem__, __setitem__

        """

        # refresh metadata
        if not self._cache_metadata:
            self._load_metadata()

        # check args
        check_fields(fields, self._dtype)

        # setup indexer
        indexer = OrthogonalIndexer(selection, self)

        return self._get_selection(indexer=indexer, out=out, fields=fields)

    def get_coordinate_selection(self, selection, out=None, fields=None):
        """Retrieve a selection of individual items, by providing the indices
        (coordinates) for each selected item.

        Parameters
        ----------
        selection : tuple
            An integer (coordinate) array for each dimension of the array.
        out : ndarray, optional
            If given, load the selected data directly into this array.
        fields : str or sequence of str, optional
            For arrays with a structured dtype, one or more fields can be specified to
            extract data for.

        Returns
        -------
        out : ndarray
            A NumPy array containing the data for the requested selection.

        Examples
        --------
        Setup a 2-dimensional array::

            >>> import zarr
            >>> import numpy as np
            >>> z = zarr.array(np.arange(100).reshape(10, 10))

        Retrieve items by specifying their coordinates::

            >>> z.get_coordinate_selection(([1, 4], [1, 4]))
            array([11, 44])

        For convenience, the coordinate selection functionality is also available via the
        `vindex` property, e.g.::

            >>> z.vindex[[1, 4], [1, 4]]
            array([11, 44])

        Notes
        -----
        Coordinate indexing is also known as point selection, and is a form of vectorized
        or inner indexing.

        Slices are not supported. Coordinate arrays must be provided for all dimensions
        of the array.

        Coordinate arrays may be multidimensional, in which case the output array will
        also be multidimensional. Coordinate arrays are broadcast against each other
        before being applied. The shape of the output will be the same as the shape of
        each coordinate array after broadcasting.

        See Also
        --------
        get_basic_selection, set_basic_selection, get_mask_selection, set_mask_selection,
        get_orthogonal_selection, set_orthogonal_selection, set_coordinate_selection,
        vindex, oindex, __getitem__, __setitem__

        """

        # refresh metadata
        if not self._cache_metadata:
            self._load_metadata()

        # check args
        check_fields(fields, self._dtype)

        # setup indexer
        indexer = CoordinateIndexer(selection, self)

        # handle output - need to flatten
        if out is not None:
            out = out.reshape(-1)

        out = self._get_selection(indexer=indexer, out=out, fields=fields)

        # restore shape
        out = out.reshape(indexer.sel_shape)

        return out

    def get_mask_selection(self, selection, out=None, fields=None):
        """Retrieve a selection of individual items, by providing a Boolean array of the
        same shape as the array against which the selection is being made, where True
        values indicate a selected item.

        Parameters
        ----------
        selection : ndarray, bool
            A Boolean array of the same shape as the array against which the selection is
            being made.
        out : ndarray, optional
            If given, load the selected data directly into this array.
        fields : str or sequence of str, optional
            For arrays with a structured dtype, one or more fields can be specified to
            extract data for.

        Returns
        -------
        out : ndarray
            A NumPy array containing the data for the requested selection.

        Examples
        --------
        Setup a 2-dimensional array::

            >>> import zarr
            >>> import numpy as np
            >>> z = zarr.array(np.arange(100).reshape(10, 10))

        Retrieve items by specifying a mask::

            >>> sel = np.zeros_like(z, dtype=bool)
            >>> sel[1, 1] = True
            >>> sel[4, 4] = True
            >>> z.get_mask_selection(sel)
            array([11, 44])

        For convenience, the mask selection functionality is also available via the
        `vindex` property, e.g.::

            >>> z.vindex[sel]
            array([11, 44])

        Notes
        -----
        Mask indexing is a form of vectorized or inner indexing, and is equivalent to
        coordinate indexing. Internally the mask array is converted to coordinate
        arrays by calling `np.nonzero`.

        See Also
        --------
        get_basic_selection, set_basic_selection, set_mask_selection,
        get_orthogonal_selection, set_orthogonal_selection, get_coordinate_selection,
        set_coordinate_selection, vindex, oindex, __getitem__, __setitem__

        """

        # refresh metadata
        if not self._cache_metadata:
            self._load_metadata()

        # check args
        check_fields(fields, self._dtype)

        # setup indexer
        indexer = MaskIndexer(selection, self)

        return self._get_selection(indexer=indexer, out=out, fields=fields)

    def _get_selection(self, indexer, out=None, fields=None):

        # We iterate over all chunks which overlap the selection and thus contain data
        # that needs to be extracted. Each chunk is processed in turn, extracting the
        # necessary data and storing into the correct location in the output array.

        # N.B., it is an important optimisation that we only visit chunks which overlap
        # the selection. This minimises the number of iterations in the main for loop.

        # check fields are sensible
        out_dtype = check_fields(fields, self._dtype)

        # determine output shape
        out_shape = indexer.shape

        # setup output array
        if out is None:
            out = np.empty_like(self._meta_array, shape=out_shape,
                                dtype=out_dtype, order=self._order)
        else:
            check_array_shape('out', out, out_shape)

        # iterate over chunks

        if math.prod(out_shape) > 0:
            # allow storage to get multiple items at once
            lchunk_coords, lchunk_selection, lout_selection = zip(*indexer)
            self._chunk_getitems(
                lchunk_coords, lchunk_selection, out, lout_selection,
                drop_axes=indexer.drop_axes, fields=fields
            )
        if out.shape:
            return out
        else:
            return out[()]

    def __setitem__(self, selection, value):
        """Modify data for an item or region of the array.

        Parameters
        ----------
        selection : tuple
            An integer index or slice or tuple of int/slice specifying the requested
            region for each dimension of the array.
        value : scalar or array-like
            Value to be stored into the array.

        Examples
        --------
        Setup a 1-dimensional array::

            >>> import zarr
            >>> z = zarr.zeros(100, dtype=int)

        Set all array elements to the same scalar value::

            >>> z[...] = 42
            >>> z[...]
            array([42, 42, 42, ..., 42, 42, 42])

        Set a portion of the array::

            >>> z[:10] = np.arange(10)
            >>> z[-10:] = np.arange(10)[::-1]
            >>> z[...]
            array([ 0, 1, 2, ..., 2, 1, 0])

        Setup a 2-dimensional array::

            >>> z = zarr.zeros((5, 5), dtype=int)

        Set all array elements to the same scalar value::

            >>> z[...] = 42

        Set a portion of the array::

            >>> z[0, :] = np.arange(z.shape[1])
            >>> z[:, 0] = np.arange(z.shape[0])
            >>> z[...]
            array([[ 0,  1,  2,  3,  4],
                   [ 1, 42, 42, 42, 42],
                   [ 2, 42, 42, 42, 42],
                   [ 3, 42, 42, 42, 42],
                   [ 4, 42, 42, 42, 42]])

        For arrays with a structured dtype, specific fields can be modified, e.g.::

            >>> a = np.array([(b'aaa', 1, 4.2),
            ...               (b'bbb', 2, 8.4),
            ...               (b'ccc', 3, 12.6)],
            ...              dtype=[('foo', 'S3'), ('bar', 'i4'), ('baz', 'f8')])
            >>> z = zarr.array(a)
            >>> z['foo'] = b'zzz'
            >>> z[...]
            array([(b'zzz', 1,   4.2), (b'zzz', 2,   8.4), (b'zzz', 3,  12.6)],
                  dtype=[('foo', 'S3'), ('bar', '<i4'), ('baz', '<f8')])

        Notes
        -----
        Slices with step > 1 are supported, but slices with negative step are not.

        Currently the implementation for __setitem__ is provided by
        :func:`vindex` if the indexing is pure fancy indexing (ie a
        broadcast-compatible tuple of integer array indices), or by
        :func:`set_basic_selection` otherwise.

        Effectively, this means that the following indexing modes are supported:

           - integer indexing
           - slice indexing
           - mixed slice and integer indexing
           - boolean indexing
           - fancy indexing (vectorized list of integers)

        For specific indexing options including outer indexing, see the
        methods listed under See Also.

        See Also
        --------
        get_basic_selection, set_basic_selection, get_mask_selection, set_mask_selection,
        get_coordinate_selection, set_coordinate_selection, get_orthogonal_selection,
        set_orthogonal_selection, vindex, oindex, __getitem__

        """
        fields, pure_selection = pop_fields(selection)
        if is_pure_fancy_indexing(pure_selection, self.ndim):
            self.vindex[selection] = value
        elif is_pure_orthogonal_indexing(pure_selection, self.ndim):
            self.set_orthogonal_selection(pure_selection, value, fields=fields)
        else:
            self.set_basic_selection(pure_selection, value, fields=fields)

    def set_basic_selection(self, selection, value, fields=None):
        """Modify data for an item or region of the array.

        Parameters
        ----------
        selection : tuple
            An integer index or slice or tuple of int/slice specifying the requested
            region for each dimension of the array.
        value : scalar or array-like
            Value to be stored into the array.
        fields : str or sequence of str, optional
            For arrays with a structured dtype, one or more fields can be specified to set
            data for.

        Examples
        --------
        Setup a 1-dimensional array::

            >>> import zarr
            >>> import numpy as np
            >>> z = zarr.zeros(100, dtype=int)

        Set all array elements to the same scalar value::

            >>> z.set_basic_selection(..., 42)
            >>> z[...]
            array([42, 42, 42, ..., 42, 42, 42])

        Set a portion of the array::

            >>> z.set_basic_selection(slice(10), np.arange(10))
            >>> z.set_basic_selection(slice(-10, None), np.arange(10)[::-1])
            >>> z[...]
            array([ 0, 1, 2, ..., 2, 1, 0])

        Setup a 2-dimensional array::

            >>> z = zarr.zeros((5, 5), dtype=int)

        Set all array elements to the same scalar value::

            >>> z.set_basic_selection(..., 42)

        Set a portion of the array::

            >>> z.set_basic_selection((0, slice(None)), np.arange(z.shape[1]))
            >>> z.set_basic_selection((slice(None), 0), np.arange(z.shape[0]))
            >>> z[...]
            array([[ 0,  1,  2,  3,  4],
                   [ 1, 42, 42, 42, 42],
                   [ 2, 42, 42, 42, 42],
                   [ 3, 42, 42, 42, 42],
                   [ 4, 42, 42, 42, 42]])

        For arrays with a structured dtype, the `fields` parameter can be used to set
        data for a specific field, e.g.::

            >>> a = np.array([(b'aaa', 1, 4.2),
            ...               (b'bbb', 2, 8.4),
            ...               (b'ccc', 3, 12.6)],
            ...              dtype=[('foo', 'S3'), ('bar', 'i4'), ('baz', 'f8')])
            >>> z = zarr.array(a)
            >>> z.set_basic_selection(slice(0, 2), b'zzz', fields='foo')
            >>> z[:]
            array([(b'zzz', 1,   4.2), (b'zzz', 2,   8.4), (b'ccc', 3,  12.6)],
                  dtype=[('foo', 'S3'), ('bar', '<i4'), ('baz', '<f8')])

        Notes
        -----
        This method provides the underlying implementation for modifying data via square
        bracket notation, see :func:`__setitem__` for equivalent examples using the
        alternative notation.

        See Also
        --------
        get_basic_selection, get_mask_selection, set_mask_selection,
        get_coordinate_selection, set_coordinate_selection, get_orthogonal_selection,
        set_orthogonal_selection, vindex, oindex, __getitem__, __setitem__

        """

        # guard conditions
        if self._read_only:
            raise ReadOnlyError()

        # refresh metadata
        if not self._cache_metadata:
            self._load_metadata_nosync()

        # handle zero-dimensional arrays
        if self._shape == ():
            return self._set_basic_selection_zd(selection, value, fields=fields)
        else:
            return self._set_basic_selection_nd(selection, value, fields=fields)

    def set_orthogonal_selection(self, selection, value, fields=None):
        """Modify data via a selection for each dimension of the array.

        Parameters
        ----------
        selection : tuple
            A selection for each dimension of the array. May be any combination of int,
            slice, integer array or Boolean array.
        value : scalar or array-like
            Value to be stored into the array.
        fields : str or sequence of str, optional
            For arrays with a structured dtype, one or more fields can be specified to set
            data for.

        Examples
        --------
        Setup a 2-dimensional array::

            >>> import zarr
            >>> import numpy as np
            >>> z = zarr.zeros((5, 5), dtype=int)

        Set data for a selection of rows::

            >>> z.set_orthogonal_selection(([1, 4], slice(None)), 1)
            >>> z[...]
            array([[0, 0, 0, 0, 0],
                   [1, 1, 1, 1, 1],
                   [0, 0, 0, 0, 0],
                   [0, 0, 0, 0, 0],
                   [1, 1, 1, 1, 1]])

        Set data for a selection of columns::

            >>> z.set_orthogonal_selection((slice(None), [1, 4]), 2)
            >>> z[...]
            array([[0, 2, 0, 0, 2],
                   [1, 2, 1, 1, 2],
                   [0, 2, 0, 0, 2],
                   [0, 2, 0, 0, 2],
                   [1, 2, 1, 1, 2]])

        Set data for a selection of rows and columns::

            >>> z.set_orthogonal_selection(([1, 4], [1, 4]), 3)
            >>> z[...]
            array([[0, 2, 0, 0, 2],
                   [1, 3, 1, 1, 3],
                   [0, 2, 0, 0, 2],
                   [0, 2, 0, 0, 2],
                   [1, 3, 1, 1, 3]])

        For convenience, this functionality is also available via the `oindex` property.
        E.g.::

            >>> z.oindex[[1, 4], [1, 4]] = 4
            >>> z[...]
            array([[0, 2, 0, 0, 2],
                   [1, 4, 1, 1, 4],
                   [0, 2, 0, 0, 2],
                   [0, 2, 0, 0, 2],
                   [1, 4, 1, 1, 4]])

        Notes
        -----
        Orthogonal indexing is also known as outer indexing.

        Slices with step > 1 are supported, but slices with negative step are not.

        See Also
        --------
        get_basic_selection, set_basic_selection, get_mask_selection, set_mask_selection,
        get_coordinate_selection, set_coordinate_selection, get_orthogonal_selection,
        vindex, oindex, __getitem__, __setitem__

        """

        # guard conditions
        if self._read_only:
            raise ReadOnlyError()

        # refresh metadata
        if not self._cache_metadata:
            self._load_metadata_nosync()

        # setup indexer
        indexer = OrthogonalIndexer(selection, self)

        self._set_selection(indexer, value, fields=fields)

    def set_coordinate_selection(self, selection, value, fields=None):
        """Modify a selection of individual items, by providing the indices (coordinates)
        for each item to be modified.

        Parameters
        ----------
        selection : tuple
            An integer (coordinate) array for each dimension of the array.
        value : scalar or array-like
            Value to be stored into the array.
        fields : str or sequence of str, optional
            For arrays with a structured dtype, one or more fields can be specified to set
            data for.

        Examples
        --------
        Setup a 2-dimensional array::

            >>> import zarr
            >>> import numpy as np
            >>> z = zarr.zeros((5, 5), dtype=int)

        Set data for a selection of items::

            >>> z.set_coordinate_selection(([1, 4], [1, 4]), 1)
            >>> z[...]
            array([[0, 0, 0, 0, 0],
                   [0, 1, 0, 0, 0],
                   [0, 0, 0, 0, 0],
                   [0, 0, 0, 0, 0],
                   [0, 0, 0, 0, 1]])

        For convenience, this functionality is also available via the `vindex` property.
        E.g.::

            >>> z.vindex[[1, 4], [1, 4]] = 2
            >>> z[...]
            array([[0, 0, 0, 0, 0],
                   [0, 2, 0, 0, 0],
                   [0, 0, 0, 0, 0],
                   [0, 0, 0, 0, 0],
                   [0, 0, 0, 0, 2]])

        Notes
        -----
        Coordinate indexing is also known as point selection, and is a form of vectorized
        or inner indexing.

        Slices are not supported. Coordinate arrays must be provided for all dimensions
        of the array.

        See Also
        --------
        get_basic_selection, set_basic_selection, get_mask_selection, set_mask_selection,
        get_orthogonal_selection, set_orthogonal_selection, get_coordinate_selection,
        vindex, oindex, __getitem__, __setitem__

        """

        # guard conditions
        if self._read_only:
            raise ReadOnlyError()

        # refresh metadata
        if not self._cache_metadata:
            self._load_metadata_nosync()

        # setup indexer
        indexer = CoordinateIndexer(selection, self)

        # handle value - need ndarray-like flatten value
        if not is_scalar(value, self._dtype):
            try:
                value = ensure_ndarray_like(value)
            except TypeError:
                # Handle types like `list` or `tuple`
                value = np.array(value, like=self._meta_array)
        if hasattr(value, 'shape') and len(value.shape) > 1:
            value = value.reshape(-1)

        self._set_selection(indexer, value, fields=fields)

    def set_mask_selection(self, selection, value, fields=None):
        """Modify a selection of individual items, by providing a Boolean array of the
        same shape as the array against which the selection is being made, where True
        values indicate a selected item.

        Parameters
        ----------
        selection : ndarray, bool
            A Boolean array of the same shape as the array against which the selection is
            being made.
        value : scalar or array-like
            Value to be stored into the array.
        fields : str or sequence of str, optional
            For arrays with a structured dtype, one or more fields can be specified to set
            data for.

        Examples
        --------
        Setup a 2-dimensional array::

            >>> import zarr
            >>> import numpy as np
            >>> z = zarr.zeros((5, 5), dtype=int)

        Set data for a selection of items::

            >>> sel = np.zeros_like(z, dtype=bool)
            >>> sel[1, 1] = True
            >>> sel[4, 4] = True
            >>> z.set_mask_selection(sel, 1)
            >>> z[...]
            array([[0, 0, 0, 0, 0],
                   [0, 1, 0, 0, 0],
                   [0, 0, 0, 0, 0],
                   [0, 0, 0, 0, 0],
                   [0, 0, 0, 0, 1]])

        For convenience, this functionality is also available via the `vindex` property.
        E.g.::

            >>> z.vindex[sel] = 2
            >>> z[...]
            array([[0, 0, 0, 0, 0],
                   [0, 2, 0, 0, 0],
                   [0, 0, 0, 0, 0],
                   [0, 0, 0, 0, 0],
                   [0, 0, 0, 0, 2]])

        Notes
        -----
        Mask indexing is a form of vectorized or inner indexing, and is equivalent to
        coordinate indexing. Internally the mask array is converted to coordinate
        arrays by calling `np.nonzero`.

        See Also
        --------
        get_basic_selection, set_basic_selection, get_mask_selection,
        get_orthogonal_selection, set_orthogonal_selection, get_coordinate_selection,
        set_coordinate_selection, vindex, oindex, __getitem__, __setitem__

        """

        # guard conditions
        if self._read_only:
            raise ReadOnlyError()

        # refresh metadata
        if not self._cache_metadata:
            self._load_metadata_nosync()

        # setup indexer
        indexer = MaskIndexer(selection, self)

        self._set_selection(indexer, value, fields=fields)

    def _set_basic_selection_zd(self, selection, value, fields=None):
        # special case __setitem__ for zero-dimensional array

        # check selection is valid
        selection = ensure_tuple(selection)
        if selection not in ((), (Ellipsis,)):
            err_too_many_indices(selection, self._shape)

        # check fields
        check_fields(fields, self._dtype)
        fields = check_no_multi_fields(fields)

        # obtain key for chunk
        ckey = self._chunk_key((0,))

        # setup chunk
        try:
            # obtain compressed data for chunk
            cdata = self.chunk_store[ckey]

        except KeyError:
            # chunk not initialized
            chunk = np.zeros_like(self._meta_array, shape=(), dtype=self._dtype)
            if self._fill_value is not None:
                chunk.fill(self._fill_value)

        else:
            # decode chunk
            chunk = self._decode_chunk(cdata).copy()

        # set value
        if fields:
            chunk[fields][selection] = value
        else:
            chunk[selection] = value

        # remove chunk if write_empty_chunks is false and it only contains the fill value
        if (not self.write_empty_chunks) and all_equal(self.fill_value, chunk):
            try:
                del self.chunk_store[ckey]
                return
            except Exception:  # pragma: no cover
                # deleting failed, fallback to overwriting
                pass
        else:
            # encode and store
            cdata = self._encode_chunk(chunk)
            self.chunk_store[ckey] = cdata

    def _set_basic_selection_nd(self, selection, value, fields=None):
        # implementation of __setitem__ for array with at least one dimension

        # setup indexer
        indexer = BasicIndexer(selection, self)

        self._set_selection(indexer, value, fields=fields)

    def _set_selection(self, indexer, value, fields=None):

        # We iterate over all chunks which overlap the selection and thus contain data
        # that needs to be replaced. Each chunk is processed in turn, extracting the
        # necessary data from the value array and storing into the chunk array.

        # N.B., it is an important optimisation that we only visit chunks which overlap
        # the selection. This minimises the number of iterations in the main for loop.

        # check fields are sensible
        check_fields(fields, self._dtype)
        fields = check_no_multi_fields(fields)

        # determine indices of chunks overlapping the selection
        sel_shape = indexer.shape

        # check value shape
        if sel_shape == ():
            # setting a single item
            pass
        elif is_scalar(value, self._dtype):
            # setting a scalar value
            pass
        else:
            if not hasattr(value, 'shape'):
                value = np.asanyarray(value, like=self._meta_array)
            check_array_shape('value', value, sel_shape)

        # iterate over chunks in range
        if not hasattr(self.chunk_store, "setitems") or self._synchronizer is not None \
           or any(map(lambda x: x == 0, self.shape)):
            # iterative approach
            for chunk_coords, chunk_selection, out_selection in indexer:

                # extract data to store
                if sel_shape == ():
                    chunk_value = value
                elif is_scalar(value, self._dtype):
                    chunk_value = value
                else:
                    chunk_value = value[out_selection]
                    # handle missing singleton dimensions
                    if indexer.drop_axes:
                        item = [slice(None)] * self.ndim
                        for a in indexer.drop_axes:
                            item[a] = np.newaxis
                        item = tuple(item)
                        chunk_value = chunk_value[item]

                # put data
                self._chunk_setitem(chunk_coords, chunk_selection, chunk_value, fields=fields)
        else:
            lchunk_coords, lchunk_selection, lout_selection = zip(*indexer)
            chunk_values = []
            for out_selection in lout_selection:
                if sel_shape == ():
                    chunk_values.append(value)
                elif is_scalar(value, self._dtype):
                    chunk_values.append(value)
                else:
                    cv = value[out_selection]
                    # handle missing singleton dimensions
                    if indexer.drop_axes:  # pragma: no cover
                        item = [slice(None)] * self.ndim
                        for a in indexer.drop_axes:
                            item[a] = np.newaxis
                        item = tuple(item)
                        cv = chunk_value[item]
                    chunk_values.append(cv)

            self._chunk_setitems(lchunk_coords, lchunk_selection, chunk_values,
                                 fields=fields)

    def _process_chunk(
        self,
        out,
        cdata,
        chunk_selection,
        drop_axes,
        out_is_ndarray,
        fields,
        out_selection,
        partial_read_decode=False,
    ):
        """Take binary data from storage and fill output array"""
        if (out_is_ndarray and
                not fields and
                is_contiguous_selection(out_selection) and
                is_total_slice(chunk_selection, self._chunks) and
                not self._filters and
                self._dtype != object):

            dest = out[out_selection]
            # Assume that array-like objects that doesn't have a
            # `writeable` flag is writable.
            dest_is_writable = getattr(dest, "writeable", True)
            write_direct = (
                dest_is_writable and
                (
                    (self._order == 'C' and dest.flags.c_contiguous) or
                    (self._order == 'F' and dest.flags.f_contiguous)
                )
            )

            if write_direct:

                # optimization: we want the whole chunk, and the destination is
                # contiguous, so we can decompress directly from the chunk
                # into the destination array
                if self._compressor:
                    if isinstance(cdata, PartialReadBuffer):
                        cdata = cdata.read_full()
                    self._compressor.decode(cdata, dest)
                else:
                    if isinstance(cdata, UncompressedPartialReadBufferV3):
                        cdata = cdata.read_full()
                    chunk = ensure_ndarray_like(cdata).view(self._dtype)
                    chunk = chunk.reshape(self._chunks, order=self._order)
                    np.copyto(dest, chunk)
                return

        # decode chunk
        try:
            if partial_read_decode:
                cdata.prepare_chunk()
                # size of chunk
                tmp = np.empty_like(self._meta_array, shape=self._chunks, dtype=self.dtype)
                index_selection = PartialChunkIterator(chunk_selection, self.chunks)
                for start, nitems, partial_out_selection in index_selection:
                    expected_shape = [
                        len(
                            range(*partial_out_selection[i].indices(self.chunks[0] + 1))
                        )
                        if i < len(partial_out_selection)
                        else dim
                        for i, dim in enumerate(self.chunks)
                    ]
                    if isinstance(cdata, UncompressedPartialReadBufferV3):
                        chunk_partial = self._decode_chunk(
                            cdata.read_part(start, nitems),
                            start=start,
                            nitems=nitems,
                            expected_shape=expected_shape,
                        )
                    else:
                        cdata.read_part(start, nitems)
                        chunk_partial = self._decode_chunk(
                            cdata.buff,
                            start=start,
                            nitems=nitems,
                            expected_shape=expected_shape,
                        )
                    tmp[partial_out_selection] = chunk_partial
                out[out_selection] = tmp[chunk_selection]
                return
        except ArrayIndexError:
            cdata = cdata.read_full()
        chunk = self._decode_chunk(cdata)

        # select data from chunk
        if fields:
            chunk = chunk[fields]
        tmp = chunk[chunk_selection]
        if drop_axes:
            tmp = np.squeeze(tmp, axis=drop_axes)

        # store selected data in output
        out[out_selection] = tmp

    def _chunk_getitems(self, lchunk_coords, lchunk_selection, out, lout_selection,
                        drop_axes=None, fields=None):
        """Obtain part or whole of chunks.

        Parameters
        ----------
        chunk_coords : list of tuple of ints
            Indices of the chunks.
        chunk_selection : list of selections
            Location of region within the chunks to extract.
        out : ndarray
            Array to store result in.
        out_selection : list of selections
            Location of regions within output array to store results in.
        drop_axes : tuple of ints
            Axes to squeeze out of the chunk.
        fields
            TODO
        """

        out_is_ndarray = True
        try:
            out = ensure_ndarray_like(out)
        except TypeError:  # pragma: no cover
            out_is_ndarray = False

        # Keys to retrieve
        ckeys = [self._chunk_key(ch) for ch in lchunk_coords]

        # Check if we can do a partial read
        if (
            self._partial_decompress
            and self._compressor
            and self._compressor.codec_id == "blosc"
            and hasattr(self._compressor, "decode_partial")
            and not fields
            and self.dtype != object
            and hasattr(self.chunk_store, "getitems")
        ):
            partial_read_decode = True
            cdatas = {
                ckey: PartialReadBuffer(ckey, self.chunk_store)
                for ckey in ckeys
                if ckey in self.chunk_store
            }
        elif (
            self._partial_decompress
            and not self._compressor
            and not fields
            and self.dtype != object
            and hasattr(self.chunk_store, "get_partial_values")
            and self.chunk_store.supports_efficient_get_partial_values
        ):
            partial_read_decode = True
            cdatas = {
                ckey: UncompressedPartialReadBufferV3(
                    ckey, self.chunk_store, itemsize=self.itemsize
                )
                for ckey in ckeys
                if ckey in self.chunk_store
            }
        elif hasattr(self.chunk_store, "get_partial_values"):
            partial_read_decode = False
            values = self.chunk_store.get_partial_values([(ckey, (0, None)) for ckey in ckeys])
            cdatas = {key: value for key, value in zip(ckeys, values) if value is not None}
        else:
            partial_read_decode = False
            contexts = {}
            if not isinstance(self._meta_array, np.ndarray):
                contexts = ConstantMap(ckeys, constant=Context(meta_array=self._meta_array))
            cdatas = self.chunk_store.getitems(ckeys, contexts=contexts)

        for ckey, chunk_select, out_select in zip(ckeys, lchunk_selection, lout_selection):
            if ckey in cdatas:
                self._process_chunk(
                    out,
                    cdatas[ckey],
                    chunk_select,
                    drop_axes,
                    out_is_ndarray,
                    fields,
                    out_select,
                    partial_read_decode=partial_read_decode,
                )
            else:
                # check exception type
                if self._fill_value is not None:
                    if fields:
                        fill_value = self._fill_value[fields]
                    else:
                        fill_value = self._fill_value
                    out[out_select] = fill_value

    def _chunk_setitems(self, lchunk_coords, lchunk_selection, values, fields=None):
        ckeys = map(self._chunk_key, lchunk_coords)
        cdatas = {key: self._process_for_setitem(key, sel, val, fields=fields)
                  for key, sel, val in zip(ckeys, lchunk_selection, values)}
        to_store = {}
        if not self.write_empty_chunks:
            empty_chunks = {k: v for k, v in cdatas.items() if all_equal(self.fill_value, v)}
            self._chunk_delitems(empty_chunks.keys())
            nonempty_keys = cdatas.keys() - empty_chunks.keys()
            to_store = {k: self._encode_chunk(cdatas[k]) for k in nonempty_keys}
        else:
            to_store = {k: self._encode_chunk(v) for k, v in cdatas.items()}
        self.chunk_store.setitems(to_store)

    def _chunk_delitems(self, ckeys):
        if hasattr(self.store, "delitems"):
            self.store.delitems(ckeys)
        else:  # pragma: no cover
            # exempting this branch from coverage as there are no extant stores
            # that will trigger this condition, but it's possible that they
            # will be developed in the future.
            tuple(map(self._chunk_delitem, ckeys))

    def _chunk_delitem(self, ckey):
        """
        Attempt to delete the value associated with ckey.
        """
        try:
            del self.chunk_store[ckey]
        except KeyError:
            pass

    def _chunk_setitem(self, chunk_coords, chunk_selection, value, fields=None):
        """Replace part or whole of a chunk.

        Parameters
        ----------
        chunk_coords : tuple of ints
            Indices of the chunk.
        chunk_selection : tuple of slices
            Location of region within the chunk.
        value : scalar or ndarray
            Value to set.

        """

        if self._synchronizer is None:
            # no synchronization
            lock = nolock
        else:
            # synchronize on the chunk
            ckey = self._chunk_key(chunk_coords)
            lock = self._synchronizer[ckey]

        with lock:
            self._chunk_setitem_nosync(chunk_coords, chunk_selection, value,
                                       fields=fields)

    def _chunk_setitem_nosync(self, chunk_coords, chunk_selection, value, fields=None):
        ckey = self._chunk_key(chunk_coords)
        cdata = self._process_for_setitem(ckey, chunk_selection, value, fields=fields)

        # attempt to delete chunk if it only contains the fill value
        if (not self.write_empty_chunks) and all_equal(self.fill_value, cdata):
            self._chunk_delitem(ckey)
        else:
            self.chunk_store[ckey] = self._encode_chunk(cdata)

    def _process_for_setitem(self, ckey, chunk_selection, value, fields=None):
        if is_total_slice(chunk_selection, self._chunks) and not fields:
            # totally replace chunk

            # optimization: we are completely replacing the chunk, so no need
            # to access the existing chunk data

            if is_scalar(value, self._dtype):

                # setup array filled with value
                chunk = np.empty_like(
                    self._meta_array, shape=self._chunks, dtype=self._dtype, order=self._order
                )
                chunk.fill(value)

            else:

                # ensure array is contiguous
                chunk = value.astype(self._dtype, order=self._order, copy=False)

        else:
            # partially replace the contents of this chunk

            try:

                # obtain compressed data for chunk
                cdata = self.chunk_store[ckey]

            except KeyError:

                # chunk not initialized
                if self._fill_value is not None:
                    chunk = np.empty_like(
                        self._meta_array, shape=self._chunks, dtype=self._dtype, order=self._order
                    )
                    chunk.fill(self._fill_value)
                elif self._dtype == object:
                    chunk = np.empty(self._chunks, dtype=self._dtype, order=self._order)
                else:
                    # N.B., use zeros here so any region beyond the array has consistent
                    # and compressible data
                    chunk = np.zeros_like(
                        self._meta_array, shape=self._chunks, dtype=self._dtype, order=self._order
                    )

            else:

                # decode chunk
                chunk = self._decode_chunk(cdata)
                if not chunk.flags.writeable:
                    chunk = chunk.copy(order='K')

            # modify
            if fields:
                # N.B., currently multi-field assignment is not supported in numpy, so
                # this only works for a single field
                chunk[fields][chunk_selection] = value
            else:
                chunk[chunk_selection] = value

        return chunk

    def _chunk_key(self, chunk_coords):
        if self._version == 3:
            # _chunk_key() corresponds to data_key(P, i, j, ...) example in the spec
            # where P = self._key_prefix,  i, j, ... = chunk_coords
            # e.g. c0/2/3 for 3d array with chunk index (0, 2, 3)
            # https://zarr-specs.readthedocs.io/en/core-protocol-v3.0-dev/protocol/core/v3.0.html#regular-grids
            return ("data/root/" + self._key_prefix +
                    "c" + self._dimension_separator.join(map(str, chunk_coords)))
        else:
            return self._key_prefix + self._dimension_separator.join(map(str, chunk_coords))

    def _decode_chunk(self, cdata, start=None, nitems=None, expected_shape=None):
        # decompress
        if self._compressor:
            # only decode requested items
            if (
                all(x is not None for x in [start, nitems])
                and self._compressor.codec_id == "blosc"
            ) and hasattr(self._compressor, "decode_partial"):
                chunk = self._compressor.decode_partial(cdata, start, nitems)
            else:
                chunk = self._compressor.decode(cdata)
        else:
            chunk = cdata

        # apply filters
        if self._filters:
            for f in reversed(self._filters):
                chunk = f.decode(chunk)

        # view as numpy array with correct dtype
        chunk = ensure_ndarray_like(chunk)
        # special case object dtype, because incorrect handling can lead to
        # segfaults and other bad things happening
        if self._dtype != object:
            chunk = chunk.view(self._dtype)
        elif chunk.dtype != object:
            # If we end up here, someone must have hacked around with the filters.
            # We cannot deal with object arrays unless there is an object
            # codec in the filter chain, i.e., a filter that converts from object
            # array to something else during encoding, and converts back to object
            # array during decoding.
            raise RuntimeError('cannot read object array without object codec')

        # ensure correct chunk shape
        chunk = chunk.reshape(-1, order='A')
        chunk = chunk.reshape(expected_shape or self._chunks, order=self._order)

        return chunk

    def _encode_chunk(self, chunk):

        # apply filters
        if self._filters:
            for f in self._filters:
                chunk = f.encode(chunk)

        # check object encoding
        if ensure_ndarray_like(chunk).dtype == object:
            raise RuntimeError('cannot write object array without object codec')

        # compress
        if self._compressor:
            cdata = self._compressor.encode(chunk)
        else:
            cdata = chunk

        # ensure in-memory data is immutable and easy to compare
        if (
            isinstance(self.chunk_store, KVStore)
            or isinstance(self._chunk_store, KVStore)
        ):
            cdata = ensure_bytes(cdata)

        return cdata

    def __repr__(self):
        t = type(self)
        r = '<{}.{}'.format(t.__module__, t.__name__)
        if self.name:
            r += ' %r' % self.name
        r += ' %s' % str(self.shape)
        r += ' %s' % self.dtype
        if self._read_only:
            r += ' read-only'
        r += '>'
        return r

    @property
    def info(self):
        """Report some diagnostic information about the array.

        Examples
        --------
        >>> import zarr
        >>> z = zarr.zeros(1000000, chunks=100000, dtype='i4')
        >>> z.info
        Type               : zarr.core.Array
        Data type          : int32
        Shape              : (1000000,)
        Chunk shape        : (100000,)
        Order              : C
        Read-only          : False
        Compressor         : Blosc(cname='lz4', clevel=5, shuffle=SHUFFLE, blocksize=0)
        Store type         : zarr.storage.KVStore
        No. bytes          : 4000000 (3.8M)
        No. bytes stored   : 320
        Storage ratio      : 12500.0
        Chunks initialized : 0/10

        """
        return self._info_reporter

    def info_items(self):
        return self._synchronized_op(self._info_items_nosync)

    def _info_items_nosync(self):

        def typestr(o):
            return '{}.{}'.format(type(o).__module__, type(o).__name__)

        def bytestr(n):
            if n > 2**10:
                return '{} ({})'.format(n, human_readable_size(n))
            else:
                return str(n)

        items = []

        # basic info
        if self.name is not None:
            items += [('Name', self.name)]
        items += [
            ('Type', typestr(self)),
            ('Data type', '%s' % self.dtype),
            ('Shape', str(self.shape)),
            ('Chunk shape', str(self.chunks)),
            ('Order', self.order),
            ('Read-only', str(self.read_only)),
        ]

        # filters
        if self.filters:
            for i, f in enumerate(self.filters):
                items += [('Filter [%s]' % i, repr(f))]

        # compressor
        items += [('Compressor', repr(self.compressor))]

        # synchronizer
        if self._synchronizer is not None:
            items += [('Synchronizer type', typestr(self._synchronizer))]

        # storage info
        items += [('Store type', typestr(self._store))]
        if self._chunk_store is not None:
            items += [('Chunk store type', typestr(self._chunk_store))]
        items += [('No. bytes', bytestr(self.nbytes))]
        if self.nbytes_stored > 0:
            items += [
                ('No. bytes stored', bytestr(self.nbytes_stored)),
                ('Storage ratio', '%.1f' % (self.nbytes / self.nbytes_stored)),
            ]
        items += [
            ('Chunks initialized', '{}/{}'.format(self.nchunks_initialized, self.nchunks))
        ]

        return items

    def digest(self, hashname="sha1"):
        """
        Compute a checksum for the data. Default uses sha1 for speed.

        Examples
        --------
        >>> import binascii
        >>> import zarr
        >>> z = zarr.empty(shape=(10000, 10000), chunks=(1000, 1000))
        >>> binascii.hexlify(z.digest())
        b'041f90bc7a571452af4f850a8ca2c6cddfa8a1ac'
        >>> z = zarr.zeros(shape=(10000, 10000), chunks=(1000, 1000))
        >>> binascii.hexlify(z.digest())
        b'7162d416d26a68063b66ed1f30e0a866e4abed60'
        >>> z = zarr.zeros(shape=(10000, 10000), dtype="u1", chunks=(1000, 1000))
        >>> binascii.hexlify(z.digest())
        b'cb387af37410ae5a3222e893cf3373e4e4f22816'
        """

        h = hashlib.new(hashname)

        for i in itertools.product(*[range(s) for s in self.cdata_shape]):
            h.update(self.chunk_store.get(self._chunk_key(i), b""))

        mkey = _prefix_to_array_key(self._store, self._key_prefix)
        h.update(self.store.get(mkey, b""))

        h.update(self.store.get(self.attrs.key, b""))

        checksum = h.digest()

        return checksum

    def hexdigest(self, hashname="sha1"):
        """
        Compute a checksum for the data. Default uses sha1 for speed.

        Examples
        --------
        >>> import zarr
        >>> z = zarr.empty(shape=(10000, 10000), chunks=(1000, 1000))
        >>> z.hexdigest()
        '041f90bc7a571452af4f850a8ca2c6cddfa8a1ac'
        >>> z = zarr.zeros(shape=(10000, 10000), chunks=(1000, 1000))
        >>> z.hexdigest()
        '7162d416d26a68063b66ed1f30e0a866e4abed60'
        >>> z = zarr.zeros(shape=(10000, 10000), dtype="u1", chunks=(1000, 1000))
        >>> z.hexdigest()
        'cb387af37410ae5a3222e893cf3373e4e4f22816'
        """

        checksum = binascii.hexlify(self.digest(hashname=hashname))

        # This is a bytes object on Python 3 and we want a str.
        if type(checksum) is not str:
            checksum = checksum.decode('utf8')

        return checksum

    def __getstate__(self):
        return {
            "store": self._store,
            "path": self._path,
            "read_only": self._read_only,
            "chunk_store": self._chunk_store,
            "synchronizer": self._synchronizer,
            "cache_metadata": self._cache_metadata,
            "cache_attrs": self._attrs.cache,
            "partial_decompress": self._partial_decompress,
            "write_empty_chunks": self._write_empty_chunks,
            "zarr_version": self._version,
            "meta_array": self._meta_array,
        }

    def __setstate__(self, state):
        self.__init__(**state)

    def _synchronized_op(self, f, *args, **kwargs):

        if self._synchronizer is None:
            # no synchronization
            lock = nolock

        else:
            # synchronize on the array
            mkey = _prefix_to_array_key(self._store, self._key_prefix)
            lock = self._synchronizer[mkey]

        with lock:
            self._refresh_metadata_nosync()
            result = f(*args, **kwargs)

        return result

    def _write_op(self, f, *args, **kwargs):

        # guard condition
        if self._read_only:
            raise ReadOnlyError()

        return self._synchronized_op(f, *args, **kwargs)

    def resize(self, *args):
        """Change the shape of the array by growing or shrinking one or more
        dimensions.

        Examples
        --------
        >>> import zarr
        >>> z = zarr.zeros(shape=(10000, 10000), chunks=(1000, 1000))
        >>> z.shape
        (10000, 10000)
        >>> z.resize(20000, 10000)
        >>> z.shape
        (20000, 10000)
        >>> z.resize(30000, 1000)
        >>> z.shape
        (30000, 1000)

        Notes
        -----
        When resizing an array, the data are not rearranged in any way.

        If one or more dimensions are shrunk, any chunks falling outside the
        new array shape will be deleted from the underlying store.
        However, it is noteworthy that the chunks partially falling inside the new array
        (i.e. boundary chunks) will remain intact, and therefore,
        the data falling outside the new array but inside the boundary chunks
        would be restored by a subsequent resize operation that grows the array size.

        """

        return self._write_op(self._resize_nosync, *args)

    def _resize_nosync(self, *args):

        # normalize new shape argument
        old_shape = self._shape
        new_shape = normalize_resize_args(old_shape, *args)
        old_cdata_shape = self._cdata_shape

        # update metadata
        self._shape = new_shape
        self._flush_metadata_nosync()

        # determine the new number and arrangement of chunks
        chunks = self._chunks
        new_cdata_shape = tuple(math.ceil(s / c)
                                for s, c in zip(new_shape, chunks))

        # remove any chunks not within range
        #   The idea is that, along each dimension,
        #     only find and remove the chunk slices that exist in 'old' but not 'new' data.
        #   Note that a mutable list ('old_cdata_shape_working_list') is introduced here
        #     to dynamically adjust the number of chunks along the already-processed dimensions
        #     in order to avoid duplicate chunk removal.
        chunk_store = self.chunk_store
        old_cdata_shape_working_list = list(old_cdata_shape)
        for idx_cdata, (val_old_cdata, val_new_cdata) in enumerate(
            zip(old_cdata_shape_working_list, new_cdata_shape)
        ):
            for cidx in itertools.product(
                *[
                    range(n_new, n_old) if (idx == idx_cdata) else range(n_old)
                    for idx, (n_old, n_new) in enumerate(
                        zip(old_cdata_shape_working_list, new_cdata_shape)
                    )
                ]
            ):
                key = self._chunk_key(cidx)
                try:
                    del chunk_store[key]
                except KeyError:
                    # chunk not initialized
                    pass
            old_cdata_shape_working_list[idx_cdata] = min(val_old_cdata, val_new_cdata)

    def append(self, data, axis=0):
        """Append `data` to `axis`.

        Parameters
        ----------
        data : array-like
            Data to be appended.
        axis : int
            Axis along which to append.

        Returns
        -------
        new_shape : tuple

        Notes
        -----
        The size of all dimensions other than `axis` must match between this
        array and `data`.

        Examples
        --------
        >>> import numpy as np
        >>> import zarr
        >>> a = np.arange(10000000, dtype='i4').reshape(10000, 1000)
        >>> z = zarr.array(a, chunks=(1000, 100))
        >>> z.shape
        (10000, 1000)
        >>> z.append(a)
        (20000, 1000)
        >>> z.append(np.vstack([a, a]), axis=1)
        (20000, 2000)
        >>> z.shape
        (20000, 2000)

        """
        return self._write_op(self._append_nosync, data, axis=axis)

    def _append_nosync(self, data, axis=0):

        # ensure data is array-like
        if not hasattr(data, 'shape'):
            data = np.asanyarray(data, like=self._meta_array)

        # ensure shapes are compatible for non-append dimensions
        self_shape_preserved = tuple(s for i, s in enumerate(self._shape)
                                     if i != axis)
        data_shape_preserved = tuple(s for i, s in enumerate(data.shape)
                                     if i != axis)
        if self_shape_preserved != data_shape_preserved:
            raise ValueError('shape of data to append is not compatible with the array; '
                             'all dimensions must match except for the dimension being '
                             'appended')

        # remember old shape
        old_shape = self._shape

        # determine new shape
        new_shape = tuple(
            self._shape[i] if i != axis else self._shape[i] + data.shape[i]
            for i in range(len(self._shape))
        )

        # resize
        self._resize_nosync(new_shape)

        # store data
        # noinspection PyTypeChecker
        append_selection = tuple(
            slice(None) if i != axis else slice(old_shape[i], new_shape[i])
            for i in range(len(self._shape))
        )
        self[append_selection] = data

        return new_shape

    def view(self, shape=None, chunks=None, dtype=None,
             fill_value=None, filters=None, read_only=None,
             synchronizer=None):
        """Return an array sharing the same data.

        Parameters
        ----------
        shape : int or tuple of ints
            Array shape.
        chunks : int or tuple of ints, optional
            Chunk shape.
        dtype : string or dtype, optional
            NumPy dtype.
        fill_value : object
            Default value to use for uninitialized portions of the array.
        filters : sequence, optional
            Sequence of filters to use to encode chunk data prior to
            compression.
        read_only : bool, optional
            True if array should be protected against modification.
        synchronizer : object, optional
            Array synchronizer.

        Notes
        -----
        WARNING: This is an experimental feature and should be used with care.
        There are plenty of ways to generate errors and/or cause data
        corruption.

        Examples
        --------

        Bypass filters:

            >>> import zarr
            >>> import numpy as np
            >>> np.random.seed(42)
            >>> labels = ['female', 'male']
            >>> data = np.random.choice(labels, size=10000)
            >>> filters = [zarr.Categorize(labels=labels,
            ...                            dtype=data.dtype,
            ...                            astype='u1')]
            >>> a = zarr.array(data, chunks=1000, filters=filters)
            >>> a[:]
            array(['female', 'male', 'female', ..., 'male', 'male', 'female'],
                  dtype='<U6')
            >>> v = a.view(dtype='u1', filters=[])
            >>> v.is_view
            True
            >>> v[:]
            array([1, 2, 1, ..., 2, 2, 1], dtype=uint8)

        Views can be used to modify data:

            >>> x = v[:]
            >>> x.sort()
            >>> v[:] = x
            >>> v[:]
            array([1, 1, 1, ..., 2, 2, 2], dtype=uint8)
            >>> a[:]
            array(['female', 'female', 'female', ..., 'male', 'male', 'male'],
                  dtype='<U6')

        View as a different dtype with the same item size:

            >>> data = np.random.randint(0, 2, size=10000, dtype='u1')
            >>> a = zarr.array(data, chunks=1000)
            >>> a[:]
            array([0, 0, 1, ..., 1, 0, 0], dtype=uint8)
            >>> v = a.view(dtype=bool)
            >>> v[:]
            array([False, False,  True, ...,  True, False, False])
            >>> np.all(a[:].view(dtype=bool) == v[:])
            True

        An array can be viewed with a dtype with a different item size, however
        some care is needed to adjust the shape and chunk shape so that chunk
        data is interpreted correctly:

            >>> data = np.arange(10000, dtype='u2')
            >>> a = zarr.array(data, chunks=1000)
            >>> a[:10]
            array([0, 1, 2, 3, 4, 5, 6, 7, 8, 9], dtype=uint16)
            >>> v = a.view(dtype='u1', shape=20000, chunks=2000)
            >>> v[:10]
            array([0, 0, 1, 0, 2, 0, 3, 0, 4, 0], dtype=uint8)
            >>> np.all(a[:].view('u1') == v[:])
            True

        Change fill value for uninitialized chunks:

            >>> a = zarr.full(10000, chunks=1000, fill_value=-1, dtype='i1')
            >>> a[:]
            array([-1, -1, -1, ..., -1, -1, -1], dtype=int8)
            >>> v = a.view(fill_value=42)
            >>> v[:]
            array([42, 42, 42, ..., 42, 42, 42], dtype=int8)

        Note that resizing or appending to views is not permitted:

            >>> a = zarr.empty(10000)
            >>> v = a.view()
            >>> try:
            ...     v.resize(20000)
            ... except PermissionError as e:
            ...     print(e)
            operation not permitted for views

        """

        store = self._store
        chunk_store = self._chunk_store
        path = self._path
        if read_only is None:
            read_only = self._read_only
        if synchronizer is None:
            synchronizer = self._synchronizer
        a = Array(store=store, path=path, chunk_store=chunk_store, read_only=read_only,
                  synchronizer=synchronizer, cache_metadata=True, zarr_version=self._version)
        a._is_view = True

        # allow override of some properties
        if dtype is None:
            dtype = self._dtype
        else:
            dtype = np.dtype(dtype)
            a._dtype = dtype
        if shape is None:
            shape = self._shape
        else:
            shape = normalize_shape(shape)
            a._shape = shape
        if chunks is not None:
            chunks = normalize_chunks(chunks, shape, dtype.itemsize)
            a._chunks = chunks
        if fill_value is not None:
            a._fill_value = fill_value
        if filters is not None:
            a._filters = filters

        return a

    def astype(self, dtype):
        """Returns a view that does on the fly type conversion of the underlying data.

        Parameters
        ----------
        dtype : string or dtype
            NumPy dtype.

        Notes
        -----
        This method returns a new Array object which is a view on the same
        underlying chunk data. Modifying any data via the view is currently
        not permitted and will result in an error. This is an experimental
        feature and its behavior is subject to change in the future.

        See Also
        --------
        Array.view

        Examples
        --------

        >>> import zarr
        >>> import numpy as np
        >>> data = np.arange(100, dtype=np.uint8)
        >>> a = zarr.array(data, chunks=10)
        >>> a[:]
        array([ 0,  1,  2,  3,  4,  5,  6,  7,  8,  9, 10, 11, 12, 13, 14, 15,
               16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31,
               32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47,
               48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63,
               64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79,
               80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95,
               96, 97, 98, 99], dtype=uint8)
        >>> v = a.astype(np.float32)
        >>> v.is_view
        True
        >>> v[:]
        array([  0.,   1.,   2.,   3.,   4.,   5.,   6.,   7.,   8.,   9.,
                10.,  11.,  12.,  13.,  14.,  15.,  16.,  17.,  18.,  19.,
                20.,  21.,  22.,  23.,  24.,  25.,  26.,  27.,  28.,  29.,
                30.,  31.,  32.,  33.,  34.,  35.,  36.,  37.,  38.,  39.,
                40.,  41.,  42.,  43.,  44.,  45.,  46.,  47.,  48.,  49.,
                50.,  51.,  52.,  53.,  54.,  55.,  56.,  57.,  58.,  59.,
                60.,  61.,  62.,  63.,  64.,  65.,  66.,  67.,  68.,  69.,
                70.,  71.,  72.,  73.,  74.,  75.,  76.,  77.,  78.,  79.,
                80.,  81.,  82.,  83.,  84.,  85.,  86.,  87.,  88.,  89.,
                90.,  91.,  92.,  93.,  94.,  95.,  96.,  97.,  98.,  99.],
              dtype=float32)
        """

        dtype = np.dtype(dtype)

        filters = []
        if self._filters:
            filters.extend(self._filters)
        filters.insert(0, AsType(encode_dtype=self._dtype, decode_dtype=dtype))

        return self.view(filters=filters, dtype=dtype, read_only=True)