blosc2 3.7.2__tar.gz → 3.8.0__tar.gz

{blosc2-3.7.2 → blosc2-3.8.0}/.pre-commit-config.yaml

@@ -7,7 +7,7 @@ ci:
 default_stages: [pre-commit, pre-push]
 repos:
   - repo: https://github.com/pre-commit/pre-commit-hooks
-    rev: v5.0.0
+    rev: v6.0.0
     hooks:
       - id: check-yaml
       - id: end-of-file-fixer
@@ -16,7 +16,7 @@ repos:
       - id: trailing-whitespace
 
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.12.7
+    rev: v0.12.11
     hooks:
       - id: ruff-check
         args: ["--fix", "--show-fixes"]

{blosc2-3.7.2 → blosc2-3.8.0}/ANNOUNCE.rst

@@ -1,15 +1,24 @@
-Announcing Python-Blosc2 3.7.2
+Announcing Python-Blosc2 3.8.0
 ==============================
 
-This is a maintenance release where:
+This is a minor version release where we have aimed to make the first steps
+towards complying fully with the array-api standard:
 
-✅ We have updated the Blosc2 C library to 2.21.1, which fixes a regression
-  in the build system detected in Fedora and Gentoo.
-✅ We reverted signature of ``TreeStore.__init__(()`` for making benchmarks
-  to get back to normal performance.
+✅ C-Blosc2 internal library updated to latest 2.21.2.
 
-Check our new blog post about ``TreeStore`` usage and performance at:
-https://www.blosc.org/posts/new-treestore-blosc2
+✅ numexpr version requirement pushed to 2.12.1 to incorporate
+``isnan``, ``isfinite``, ``isinf`` functions.
+
+✅ Indexing is now supported extensively and reasonably optimally for slices
+with negative steps, general boolean arrays and fancy indices, with both get/setitem having
+equal functionality. In PR #459 we extended the 1D fast path to general N-D,
+with consequent speedups. In PR #461 we allowed fancy indexing and general slicing
+with negative steps for set and getitem, with a memory-optimised path for setitem.
+
+✅ Various attributes and methods for the ``NDArray`` class, as well as functions, have
+been added to increase compliance with the array-api standard. In addition,
+linspace and arange functions have been made more numerically stable and now strictly
+comply even with difficult floating-point edge cases.
 
 You can think of Python-Blosc2 3.x as an extension of NumPy/numexpr that:
 

{blosc2-3.7.2 → blosc2-3.8.0}/CMakeLists.txt

@@ -50,7 +50,7 @@ else()
     include(FetchContent)
     FetchContent_Declare(blosc2
         GIT_REPOSITORY https://github.com/Blosc/c-blosc2
-        GIT_TAG 96bce728dcdbf41fd86f142ebef9e513f87a7afb  # v2.21.1
+        GIT_TAG 82a85f9d285d7ae6d141c73e84c7574bfc24a134  # v2.21.2
     )
     FetchContent_MakeAvailable(blosc2)
     include_directories("${blosc2_SOURCE_DIR}/include")

{blosc2-3.7.2 → blosc2-3.8.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: blosc2
-Version: 3.7.2
+Version: 3.8.0
 Summary: A fast & compressed ndarray library with a flexible compute engine.
 Author-Email: Blosc Development Team <blosc@blosc.org>
 Maintainer-Email: Blosc Development Team <blosc@blosc.org>
@@ -26,7 +26,7 @@ Requires-Dist: numpy>=1.26
 Requires-Dist: ndindex
 Requires-Dist: msgpack
 Requires-Dist: platformdirs
-Requires-Dist: numexpr; platform_machine != "wasm32"
+Requires-Dist: numexpr>=2.12.1; platform_machine != "wasm32"
 Requires-Dist: py-cpuinfo; platform_machine != "wasm32"
 Requires-Dist: requests
 Provides-Extra: dev

{blosc2-3.7.2 → blosc2-3.8.0}/README_DEVELOPERS.md

@@ -70,3 +70,13 @@ We are using Sphinx for documentation.  You can build the documentation by execu
 [You may need to install the `pandoc` package first: https://pandoc.org/installing.html]
 
 You will find the documentation in the `../html` directory.
+
+## Array API tests compatibility
+
+You can test array API compatibility with the `array-api-tests` module.
+Use the `tests/array-api-xfails.txt` to skip the tests that are not supported
+and run pytest from the `array-api-tests` source dir like this:
+
+``` bash
+ARRAY_API_TESTS_MODULE=blosc2 pytest array_api_tests --xfails-file ${BLOSC2_DIR}/tests/array-api-xfails.txt -xs
+```

{blosc2-3.7.2 → blosc2-3.8.0}/RELEASE_NOTES.md

@@ -1,5 +1,27 @@
 # Release notes
 
+## Changes from 3.8.0 to 3.8.1
+
+XXX version-specific blurb XXX
+
+## Changes from 3.7.2 to 3.8.0
+
+* C-Blosc2 internal library updated to latest 2.21.2.
+
+* numexpr version requirement pushed to 2.12.1 to incorporate
+``isnan``, ``isfinite``, ``isinf`` functions.
+
+* Indexing is now supported extensively and reasonably optimally for slices
+with negative steps and general boolean arrays, with both get/setitem having
+equal functionality. In PR #459 we extended the 1D fast path to general N-D,
+with consequent speedups. In PR # we allowed fancy indexing and general slicing
+with negative steps for set and getitem, with a memory-optimised path for setitem.
+
+* Various attributes and methods for the ``NDArray`` class, as well as functions, have
+been added to increase compliance with the array-api standard. In addition,
+linspace and arange functions have been made more numerically stable and now strictly
+comply even with difficult floating-point edge cases.
+
 ## Changes from 3.7.1 to 3.7.2
 
 * C-Blosc2 internal library updated to latest 2.21.1.

blosc2-3.8.0/ROADMAP-TO-4.0.md

@@ -0,0 +1,23 @@
+List of desired features for a 4.0 release
+------------------------------------------
+
+* First and foremost, we would like to have at least of basic implementation of the [array API](https://data-apis.org/array-api).  This will require a lot of low-level work on the basic NDArray container to make indexing to work as close as the standard.
+
+* Have a completely specified format for the `TreeStore` and `DictStore`.  The format should allow to have containers either in memory or on disk.  Also, it should allow a sparse or contiguous storage.  The user will be able to specify these properties by following the same conventions than for NDArray objects (alas, `urlpath` and `contiguous` params).
+
+    * New `.save()` and `.to_cframe()` methods should be implemented to convert from in-memory representations to on disk and viceversa.
+    * The format for `TreeStore` and `DictStore` will initially be defined at Python level, and documented only in the Python-Blosc2 repository.  An implementation in the C library is desirable, but not mandatory at this time.
+
+* A new `Table` object should be implemented based on the `TreeStore` class (a subclass?), with a label ('table'?) in metalayers indicating that the contents of the tree can be interpreted as regular table.  As `TreeStore` is hierarchical, a subtree can also be interpreted as a `Table` if there a label in the metalayer of the subtree (or group in HDF5 parlance); that can lead to tables than can have different subtables embedded.  It is not clear yet if should impose the same number of rows for all the columns.
+
+The constructor for the `Table` object should take some parameters to specify properties:
+
+    * `columnar`: True or False. If True, every column will be stored in a different NDArray object.  If False, the columns will be stored in the same NDArray object, with a compound dtype.  In principle, one should be able to create tables that are hybrid between column and row wise, but at this point it is not clear what is the best way to do that.
+
+`Table` should support at least these methods:
+
+    * `.__getitem__()` and `.__setitem__()` so that values can be get and set.
+    * `.append()` for appending (multi-) rows of data for all columns in one go.
+    * `.__iter__()` for easy and fast iteration over rows.
+    * `.where()`: an iterator for queying with conditions that are evaluated with the internal compute engine.
+    * `.index()` for indexing a column and getting better performance in queries (desirable, but optional for 4.0).

{blosc2-3.7.2 → blosc2-3.8.0}/doc/python-blosc2.rst

@@ -6,7 +6,7 @@
 
     <p style="text-align: center; color: black; background-color: rgba(230, 169, 9, 0.65);">
         <a href="https://github.com/Blosc/python-blosc2/blob/main/RELEASE_NOTES.md"
-           style="font-size: 1.5em;">Version 3.7.1 released on 2025-08-17!</a>
+           style="font-size: 1.5em;">Version 3.7.2 released on 2025-08-19!</a>
         <span style="display: inline-block; width: 20px;"></span>
         <span style="font-family: monospace;">pip install blosc2 -U</span>
     </p>
@@ -36,7 +36,7 @@
                 <h3 style="margin-top: 10px; margin-bottom: 15px;">Top-Notch Compression</h3>
             </div>
 
-        `Combine advanced codecs and filters <https://blosc.org/docs/LEAPS-INNOV_WP7_D74_v1.pdf>`_ for efficient `lossless <https://www.blosc.org/posts/bytedelta-enhance-compression-toolset/>`_ and `lossy <https://www.blosc.org/posts/blosc2-lossy-compression/>`_ compression to reduce storage space while maintaining high performance.
+        `Combine advanced codecs and filters <https://www.blosc.org/docs/2025-EuroSciPy-Blosc2.pdf>`_ for efficient `lossless <https://www.blosc.org/posts/bytedelta-enhance-compression-toolset/>`_ and `lossy <https://www.blosc.org/posts/blosc2-lossy-compression/>`_ compression to reduce storage space while maintaining high performance.
 
     .. grid-item-card::
         :class-card: intro-card text-center no-border
@@ -48,7 +48,7 @@
                 <h3 style="margin-top: 10px; margin-bottom: 15px;">Full-Fledged NDArrays</h3>
             </div>
 
-        `NDArray objects <https://www.blosc.org/python-blosc2/getting_started/tutorials/01.ndarray-basics.html>`_ enable efficient storage and manipulation of arbitrarily large N-dimensional datasets, following the `Array API <https://data-apis.org/array-api/latest/>`_ standard, with an additional `C-API <https://www.blosc.org/posts/blosc2-ndim-intro/>`_ available.
+        `NDArray objects <https://www.blosc.org/python-blosc2/getting_started/tutorials/01.ndarray-basics.html>`_ enable efficient storage and manipulation of arbitrarily large N-dimensional datasets, following the `Array API <https://data-apis.org/array-api/latest/>`_ standard, with an additional `C API <https://www.blosc.org/posts/blosc2-ndim-intro/>`_.
 
     .. grid-item-card::
         :class-card: intro-card text-center no-border
@@ -60,7 +60,7 @@
                 <h3 style="margin-top: 10px; margin-bottom: 15px;">Compute Engine Inside</h3>
             </div>
 
-        Combines compression with high-speed computation of complex `mathematical expressions <https://ironarray.io/blog/blosc2-eval-expressions>`_ and `reductions <https://www.blosc.org/posts/ndim-reductions/>`_, while maintaining compatibility with NumPy.
+        Combines compression with high-speed computation of complex `mathematical expressions <https://ironarray.io/blog/blosc2-eval-expressions>`_ and `reductions <https://www.blosc.org/python-blosc2/getting_started/tutorials/04.reductions.html>`_, while maintaining compatibility with NumPy.
 
     .. grid-item-card::
         :class-card: intro-card text-center no-border
@@ -68,11 +68,11 @@
         .. raw:: html
 
             <div style="text-align: center;">
-                <i class="fas fa-th" style="font-size: 2em; color: #007acc; margin-bottom: 10px;"></i>
-                <h3 style="margin-top: 10px; margin-bottom: 15px;">Sparse Data Mastery</h3>
+                <i class="fas fa-sitemap" style="font-size: 2em; color: #007acc; margin-bottom: 10px;"></i>
+                <h3 style="margin-top: 10px; margin-bottom: 15px;">Hierarchical Structures</h3>
             </div>
 
-        Efficiently store and manipulate sparse data with `compact sparse constructs <https://www.blosc.org/docs/Exploring-MilkyWay-SciPy2023.pdf>`_ for optimized `performance <https://www.blosc.org/docs/Exploring-MilkyWay-SciPy2023-paper.pdf>`_.
+        Efficiently store data hierachically with the `TreeStore class <https://www.blosc.org/python-blosc2/reference/tree_store.html#blosc2.TreeStore>`_ for convenience and optimized `performance <https://www.blosc.org/posts/new-treestore-blosc2/>`_.
 
     .. grid-item-card::
         :class-card: intro-card text-center no-border
@@ -84,7 +84,7 @@
                 <h3 style="margin-top: 10px; margin-bottom: 15px;">Flexible Storage</h3>
             </div>
 
-        Access data from anywhere: read/write in `memory or disk <https://github.com/Blosc/python-blosc2/blob/main/doc/getting_started/tutorials/01.ndarray-basics.ipynb>`_, stream from `the network <https://www.blosc.org/python-blosc2/reference/c2array.html>`_, or use `memory-mapped files <https://www.blosc.org/python-blosc2/reference/storage.html#blosc2.Storage.params.mmap_mode>`_ for high-performance I/O.
+        Access data from anywhere: read/write in `memory or disk <https://www.blosc.org/docs/Exploring-MilkyWay-SciPy2023-paper.pdf>`_, stream from `the network <https://www.blosc.org/python-blosc2/reference/c2array.html>`_, or use `memory-mapped files <https://www.blosc.org/python-blosc2/reference/storage.html#blosc2.Storage.params.mmap_mode>`_ for high-performance I/O.
 
     .. grid-item-card::
         :class-card: intro-card text-center no-border

{blosc2-3.7.2 → blosc2-3.8.0}/examples/ndarray/animated_plot.py

@@ -26,7 +26,7 @@ n_frames = int(1000 * scale)  # Raise this for more frames
 dtype = np.float64  # Data type for the grid
 use_blosc2 = True  # Set to False to use NumPy arrays instead
 realize_blosc2 = False  # Set to False to skip Blosc2 realization
-make_animation = False  # Set to False to skip animation creation
+make_animation = True  # Set to False to skip animation creation
 travel_dim = 2  # Dimension to travel through (0 for X, 1 for Y, 2 for Z)
 
 # --- Coordinate creation ---

{blosc2-3.7.2 → blosc2-3.8.0}/pyproject.toml

@@ -37,11 +37,13 @@ dependencies = [
     "ndindex",
     "msgpack",
     "platformdirs",
-    "numexpr; platform_machine != 'wasm32'",
+    "numexpr>=2.12.1; platform_machine != 'wasm32'",
     "py-cpuinfo; platform_machine != 'wasm32'",
     "requests",
 ]
-version = "3.7.2"
+version = "3.8.0"
+[project.entry-points."array_api"]
+blosc2 = "blosc2"
 
 
 [project.optional-dependencies]
@@ -114,7 +116,7 @@ extend-select = [
     "SIM",
     "TC",
     "UP",
-]
+    "C901"]  # enable complexity rule
 ignore = [
     "B028",
     "PT011",
@@ -128,3 +130,7 @@ ignore = [
 
 [tool.ruff.lint.extend-per-file-ignores]
 "tests/**" = ["F841"]
+
+[tool.ruff.lint.mccabe]
+# Raise complexity from the default 10 to 13
+max-complexity = 13

{blosc2-3.7.2 → blosc2-3.8.0}/src/blosc2/__init__.py

@@ -13,6 +13,8 @@
 import platform
 from enum import Enum
 
+import numpy as np
+
 # Do the platform check once at module level
 IS_WASM = platform.machine() == "wasm32"
 # IS_WASM = True  # for testing (comment this line out for production)
@@ -23,9 +25,10 @@ Flag for WebAssembly platform.
 if not IS_WASM:
     import numexpr
 
-from .version import __version__
+from .version import __array_api_version__, __version__
 
 __version__ = __version__
+__array_api_version__ = __array_api_version__
 """
 Python-Blosc2 version.
 """
@@ -164,6 +167,94 @@ VERSION_STRING = VERSION_STRING
 """
 The C-Blosc2 version's string."""
 
+
+# For array-api compatibility
+iinfo = np.iinfo
+finfo = np.finfo
+
+# dtypes for array-api
+str_ = np.str_
+bytes_ = np.bytes_
+object_ = np.object_
+
+from numpy import (
+    bool_,
+    complex64,
+    complex128,
+    e,
+    euler_gamma,
+    float16,
+    float32,
+    float64,
+    inf,
+    int8,
+    int16,
+    int32,
+    int64,
+    nan,
+    newaxis,
+    pi,
+    uint8,
+    uint16,
+    uint32,
+    uint64,
+)
+
+DEFAULT_COMPLEX = complex128
+"""
+Default complex floating dtype."""
+
+DEFAULT_FLOAT = float64
+"""
+Default real floating dtype."""
+
+DEFAULT_INT = int64
+"""
+Default integer dtype."""
+
+DEFAULT_INDEX = int64
+"""
+Default indexing dtype."""
+
+
+class Info:
+    def __init__(self, **kwargs):
+        for key, value in kwargs.items():
+            setattr(self, key, value)
+
+
+def __array_namespace_info__() -> Info:
+    """
+    Return information about the array namespace following the Array API specification.
+    """
+
+    def _raise(exc):
+        raise exc
+
+    return Info(
+        capabilities=lambda: {
+            "boolean indexing": True,
+            "data-dependent shapes": False,
+            "max dimensions": MAX_DIM,
+        },
+        default_device=lambda: "cpu",
+        default_dtypes=lambda device=None: {
+            "real floating": DEFAULT_FLOAT,
+            "complex floating": DEFAULT_COMPLEX,
+            "integral": DEFAULT_INT,
+            "indexing": DEFAULT_INDEX,
+        }
+        if (device == "cpu" or device is None)
+        else _raise(ValueError("Only cpu devices allowed")),
+        dtypes=lambda device=None, kind=None: np.__array_namespace_info__().dtypes(kind=kind, device=device)
+        if (device == "cpu" or device is None)
+        else _raise(ValueError("Only cpu devices allowed")),
+        devices=lambda: ["cpu"],
+        name="blosc2",
+        version=__version__,
+    )
+
+
 # Public API for container module
 from .core import (
     clib_info,
@@ -237,9 +328,11 @@ from .ndarray import (
     are_partitions_aligned,
     are_partitions_behaved,
     arange,
+    broadcast_to,
     linspace,
     eye,
     asarray,
+    astype,
     indices,
     sort,
     reshape,
@@ -248,14 +341,19 @@ from .ndarray import (
     concatenate,
     expand_dims,
     empty,
+    empty_like,
     frombuffer,
     fromiter,
     get_slice_nchunks,
+    meshgrid,
     nans,
     uninit,
     zeros,
+    zeros_like,
     ones,
+    ones_like,
     full,
+    full_like,
     save,
     matmul,
     permute_dims,
@@ -278,6 +376,7 @@ from .lazyexpr import (
     get_expr_operands,
     validate_expr,
     evaluate,
+    slices_eval,
 )
 from .proxy import Proxy, ProxySource, ProxyNDSource, ProxyNDField, SimpleProxy, jit
 
@@ -328,13 +427,18 @@ from .ndarray import (
     arctan,
     arctan2,
     arctanh,
+    array_from_ffi_ptr,
     conj,
     contains,
     cos,
     cosh,
+    equal,
     exp,
     expm1,
     imag,
+    isfinite,
+    isinf,
+    isnan,
     lazywhere,
     log,
     log1p,
@@ -355,7 +459,7 @@ from .ndarray import (
     where,
 )
 
-__all__ = [
+__all__ = [  # noqa : RUF022
     # Constants
     "EXTENDED_HEADER_LENGTH",
     "MAX_BUFFERSIZE",
@@ -363,6 +467,17 @@ __all__ = [
     "MIN_HEADER_LENGTH",
     "VERSION_DATE",
     "VERSION_STRING",
+    # Default dtypes
+    "DEFAULT_COMPLEX",
+    "DEFAULT_FLOAT",
+    "DEFAULT_INDEX",
+    "DEFAULT_INT",
+    # Mathematical constants
+    "e",
+    "pi",
+    "inf",
+    "nan",
+    "newaxis",
     # Classes
     "C2Array",
     "CParams",
@@ -406,7 +521,8 @@ __all__ = [
     "are_partitions_aligned",
     "are_partitions_behaved",
     "asarray",
-    "clib_info",
+    "astypeclib_info",
+    "broadcast_to",
     "compress",
     "compress2",
     "compressor_list",
@@ -420,6 +536,9 @@ __all__ = [
     "decompress2",
     "detect_number_of_cores",
     "dparams_dflts",
+    "empty",
+    "empty_like",
+    "equal",
     "estore_from_cframe",
     "expand_dims",
     "expm1",
@@ -429,6 +548,7 @@ __all__ = [
     "frombuffer",
     "fromiter",
     "full",
+    "full_like",
     "get_blocksize",
     "get_cbuffer_sizes",
     "get_clib",
@@ -437,6 +557,9 @@ __all__ = [
     "get_expr_operands",
     "get_slice_nchunks",
     "indices",
+    "isfinite",
+    "isinf",
+    "isnan",
     "jit",
     "lazyexpr",
     "lazyudf",
@@ -451,10 +574,12 @@ __all__ = [
     "matrix_transpose",
     "max",
     "mean",
+    "meshgrid",
     "min",
     "nans",
     "ndarray_from_cframe",
     "ones",
+    "ones_like",
     "open",
     "pack",
     "pack_array",
@@ -493,4 +618,6 @@ __all__ = [
     "validate_expr",
     "var",
     "where",
+    "zeros",
+    "zeros_like",
 ]

{blosc2-3.7.2 → blosc2-3.8.0}/src/blosc2/blosc2_ext.pyx

@@ -2612,8 +2612,12 @@ cdef class NDArray:
                 mask_[i] = mask[i]
             _check_rc(b2nd_squeeze_index(self.array, mask_), "Error while squeezing array")
 
-        if self.array.shape[0] == 1 and self.ndim == 1:
-            self.array.ndim = 0
+        #this squeezes even if not asked for by mask - may have to use in future though
+        #if self.array.shape[0] == 1 and self.ndim == 1:
+        #    self.array.ndim = 0
+
+    def as_ffi_ptr(self):
+        return PyCapsule_New(self.array, <char *> "b2nd_array_t*", NULL)
 
     cdef udf_udata *_fill_udf_udata(self, func_id, inputs_id):
         cdef udf_udata *udata = <udf_udata *> malloc(sizeof(udf_udata))
@@ -2901,6 +2905,10 @@ def asarray(ndarray, chunks, blocks, **kwargs):
 
     return ndarray
 
+def array_from_ffi_ptr(array_ptr):
+    array = <b2nd_array_t *> PyCapsule_GetPointer(array_ptr, <char *> "b2nd_array_t*")
+    return blosc2.NDArray(_schunk=PyCapsule_New(array.sc, <char *> "blosc2_schunk*", NULL),
+                             _array=array_ptr)
 
 def ndarray_from_cframe(cframe, copy=False):
     cdef Py_buffer buf

{blosc2-3.7.2 → blosc2-3.8.0}/src/blosc2/c2array.py

@@ -45,13 +45,13 @@ def c2context(
     A parameter not specified or set to ``None`` will inherit the value from the
     previous context manager, defaulting to an environment variable (see
     below) if supported by that parameter.  Parameters set to an empty string
-    will not to be used in requests (with no default either).
+    will not be used in requests (without a default either).
 
     If the subscriber requires authorization for requests, you can either
     provide an `auth_token` (which you should have obtained previously from the
     subscriber), or both `username` and `password` to obtain the token by
     logging in to the subscriber.  The token will be reused until it is explicitly
-    reset or requested again in a subsequent context manager invocation.
+    reset or requested again in a later context manager invocation.
 
     Please note that this manager is reentrant but not safe for concurrent use.
 
@@ -190,6 +190,10 @@ class C2Array(blosc2.Operand):
     def __init__(self, path: str, /, urlbase: str | None = None, auth_token: str | None = None):
         """Create an instance of a remote NDArray.
 
+        Remote NDArrays can be accessed via HTTP from a Caterva2 server
+        (e.g., https://cat2.cloud). More information about Caterva2 at:
+        https://ironarray.io/caterva2.
+
         Parameters
         ----------
         path: str

{blosc2-3.7.2 → blosc2-3.8.0}/src/blosc2/core.py

@@ -1467,7 +1467,7 @@ def compute_chunks_blocks(  # noqa: C901
 
     # Return an arbitrary value for chunks and blocks when shape has any 0 dim
     if 0 in shape:
-        return (1,) * len(shape), (1,) * len(shape)
+        return shape, shape
 
     if blocks:
         if not isinstance(blocks, tuple | list):

{blosc2-3.7.2 → blosc2-3.8.0}/src/blosc2/embed_store.py

@@ -31,7 +31,7 @@ class EmbedStore:
     Parameters
     ----------
     urlpath : str or None, optional
-        Path for persistent storage. Using a ' .b2e' extension is recommended.
+        Path for persistent storage. Using a '.b2e' extension is recommended.
         If None, the embed store will be in memory only, which can be
         deserialized later using the :func:`blosc2.from_cframe` function.
     mode : str, optional