Script 'mail_helper' called by obssrc
Hello community,
here is the log from the commit of package python-h5py for openSUSE:Factory
checked in at 2023-08-23 14:58:47
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Comparing /work/SRC/openSUSE:Factory/python-h5py (Old)
and /work/SRC/openSUSE:Factory/.python-h5py.new.1766 (New)
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Package is "python-h5py"
Wed Aug 23 14:58:47 2023 rev:26 rq:1105464 version:3.9.0
Changes:
--------
--- /work/SRC/openSUSE:Factory/python-h5py/python-h5py.changes 2023-02-16
16:57:22.376334183 +0100
+++ /work/SRC/openSUSE:Factory/.python-h5py.new.1766/python-h5py.changes
2023-08-23 14:59:55.366237773 +0200
@@ -1,0 +2,47 @@
+Tue Aug 22 18:22:43 UTC 2023 - Ben Greiner <c...@bnavigator.de>
+
+- Update to 3.9.0
+ * This version of h5py requires Python 3.8 or above.
+ ## New features
+ * New out argument to read_direct_chunk() to allow passing the
+ output buffer (PR 2232).
+ * The objects from Dataset.asstr() and Dataset.astype() now
+ implement the __array__() method (PR 2269). This speeds up
+ access for functions that support it, such as np.asarray().
+ * Validate key types when creating groups and attributes, giving
+ better error messages when invalid types are used (PR 2266).
+ ## Deprecations & removals
+ * Using Dataset.astype() as a context manager has been removed,
+ after being deprecated in h5py 3.6. Read data by slicing the
+ returned object instead: dset.astype('f4')[:].
+ * Exposing HDF5 functions
+ * H5Pget_elink_acc_flags & H5Pset_elink_acc_flags as
+ h5py.h5p.PropLAID.get_elink_acc_flags() &
+ h5py.h5p.PropLAID.set_elink_acc_flags(): access the external
+ link file access traversal flags in a link access property list
+ (PR 2244).
+ * H5Zregister as h5py.h5z.register_filter(): register an HDF5
+ filter (PR 2229).
+ ## Bug fixes
+ * Group.__contains__ and Group.get now use the default link
+ access property list systematically (PR 2244).
+ * Removed various calls to the deprecated numpy.product function
+ (PR 2242 & PR 2273).
+ * Fix the IPython tab-completion integration in IPython 8.12 (PR
+ 2256).
+ * Replacing attributes with AttributeManager.create() now deletes
+ the old attributes before creating the new one, rather than
+ using a temporary name and renaming the new attribute (PR
+ 2274). This should avoid some confusing bugs affecting
+ attributes. However, failures creating an attribute are less
+ likely to leave an existing attribute of the same name in
+ place. To change an attribute value without changing its shape
+ or dtype, use modify() instead.
+ ## Building h5py
+ * When building with Parallel HDF5 support, the version of mpi4py
+ used on various Python versions is increased to 3.1.1, fixing
+ building with a newer setuptools (PR 2225).
+ * Some fixes towards compatibility with the upcoming Cython 3 (PR
+ 2247).
+
+-------------------------------------------------------------------
Old:
----
h5py-3.8.0.tar.gz
New:
----
h5py-3.9.0.tar.gz
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Other differences:
------------------
++++++ python-h5py.spec ++++++
--- /var/tmp/diff_new_pack.D8EuFm/_old 2023-08-23 14:59:56.026238953 +0200
+++ /var/tmp/diff_new_pack.D8EuFm/_new 2023-08-23 14:59:56.034238967 +0200
@@ -60,16 +60,16 @@
%endif
# /SECTION MPI DEFINITIONS
Name: %{pname}%{?my_suffix}
-Version: 3.8.0
+Version: 3.9.0
Release: 0
Summary: Python interface to the Hierarchical Data Format library
License: BSD-3-Clause
Group: Development/Libraries/Python
URL: https://github.com/h5py/h5py
Source:
https://files.pythonhosted.org/packages/source/h/h5py/h5py-%{version}.tar.gz
-BuildRequires: %{python_module Cython >= 0.29}
-BuildRequires: %{python_module devel >= 3.7}
-BuildRequires: %{python_module numpy-devel >= 1.14.5}
+BuildRequires: %{python_module Cython >= 0.29 with %python-Cython < 1}
+BuildRequires: %{python_module devel >= 3.8}
+BuildRequires: %{python_module numpy-devel >= 1.17.3}
BuildRequires: %{python_module pip}
BuildRequires: %{python_module pkgconfig}
BuildRequires: %{python_module pytest}
@@ -80,12 +80,13 @@
BuildRequires: python-rpm-macros
%requires_eq hdf5%{?my_suffix}
%requires_eq libhdf5%{?my_suffix}
-Requires: python-numpy >= 1.14.5
+Requires: python-numpy >= 1.17.3
%if %{with mpi}
BuildRequires: %{mpi_flavor}%{mpi_vers}-devel
-BuildRequires: %{python_module mpi4py >= 3.0.2}
+BuildRequires: %{python_module mpi4py >= 3.1.1 if %python-base < 3.11}
+BuildRequires: %{python_module mpi4py >= 3.1.4 if %python-base >= 3.11}
BuildRequires: %{python_module pytest-mpi}
-Requires: python-mpi4py >= 3.0.2
+Requires: python-mpi4py >= 3.1.1
%endif
%python_subpackages
@@ -119,7 +120,12 @@
%python_expand %fdupes %{buildroot}%{my_sitearch_in_expand}/h5py/
%check
-%{python_expand # Offset test fails on 32-bit
+donttest="dummytest"
+%ifarch %{ix86} %{arm}
+# overflow
+donttest="test_float_round_tripping or test_register_filter"
+%endif
+%{python_expand #
%if %{with mpi}
source %{my_bindir}/mpivars.sh
%endif
@@ -127,11 +133,7 @@
export PYTHONPATH=%{buildroot}%{my_sitearch_in_expand}
export PYTHONDONTWRITEBYTECODE=1
pytest-%{$python_bin_suffix} %{buildroot}%{my_sitearch_in_expand}/h5py/ \
-%ifarch %{ix86}
- %{?with_mpi:-k 'not test_float_round_tripping' -m 'not
mpi_skip'}%{!?with_mpi:-k 'not (TestMPI or test_float_round_tripping)'}
-%else
- %{?with_mpi:-m 'not mpi_skip'}%{!?with_mpi:-k 'not TestMPI'}
-%endif
+%{?with_mpi:-k "not ($donttest)" -m 'not mpi_skip'}%{!?with_mpi:-k "not
(TestMPI or $donttest)"}
}
%files %{python_files}
++++++ h5py-3.8.0.tar.gz -> h5py-3.9.0.tar.gz ++++++
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/.gitignore new/h5py-3.9.0/.gitignore
--- old/h5py-3.8.0/.gitignore 2022-12-21 17:13:47.000000000 +0100
+++ new/h5py-3.9.0/.gitignore 1970-01-01 01:00:00.000000000 +0100
@@ -1,40 +0,0 @@
-h5py/h5*.c
-h5py/utils.c
-h5py/_conv.c
-h5py/_proxy.c
-h5py/_objects.c
-h5py/_errors.c
-h5py/_selector.c
-h5py/config.pxi
-h5py/_hdf5.*
-h5py/*.dll
-h5py/*.so
-*.hdf5
-h5config.json
-h5py/defs.*
-build/
-*.pyc
-*.pyd
-dist/
-MANIFEST
-.DS_Store
-/docs/_build
-/docs_api/_build
-/.tox
-/h5py.egg-info
-/*.egg
-/.asv
-.eggs/
-*.so
-*~
-*.swp
-.pytest_cache
-.coverage
-.coverage_dir
-coverage.xml
-.vscode/
-.mypy_cache/
-
-# Rever
-rever/
-.idea/
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/PKG-INFO new/h5py-3.9.0/PKG-INFO
--- old/h5py-3.8.0/PKG-INFO 2023-01-23 11:21:36.018553500 +0100
+++ new/h5py-3.9.0/PKG-INFO 2023-06-20 11:06:12.781311300 +0200
@@ -1,6 +1,6 @@
Metadata-Version: 2.1
Name: h5py
-Version: 3.8.0
+Version: 3.9.0
Summary: Read and write HDF5 files from Python
Author-email: Andrew Collette <andrew.colle...@gmail.com>
Maintainer-email: Thomas Kluyver <tho...@kluyver.me.uk>, Thomas A Caswell
<tcasw...@bnl.gov>
@@ -26,7 +26,7 @@
Classifier: Topic :: Scientific/Engineering
Classifier: Topic :: Database
Classifier: Topic :: Software Development :: Libraries :: Python Modules
-Requires-Python: >=3.7
+Requires-Python: >=3.8
Description-Content-Type: text/x-rst
License-File: LICENSE
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/README.rst new/h5py-3.9.0/README.rst
--- old/h5py-3.8.0/README.rst 2023-01-20 17:24:30.000000000 +0100
+++ new/h5py-3.9.0/README.rst 2023-06-07 13:39:16.000000000 +0200
@@ -20,7 +20,7 @@
Installation
------------
-Pre-build `h5py` can either be installed via your Python Distribution (e.g.
+Pre-built `h5py` can either be installed via your Python Distribution (e.g.
`Continuum Anaconda`_, `Enthought Canopy`_) or from `PyPI`_ via `pip`_.
`h5py` is also distributed in many Linux Distributions (e.g. Ubuntu, Fedora),
and in the macOS package managers `Homebrew <https://brew.sh/>`_,
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/docs/conf.py new/h5py-3.9.0/docs/conf.py
--- old/h5py-3.8.0/docs/conf.py 2023-01-23 11:18:46.000000000 +0100
+++ new/h5py-3.9.0/docs/conf.py 2023-06-20 10:56:46.000000000 +0200
@@ -62,7 +62,7 @@
# built documents.
#
# The full version, including alpha/beta/rc tags.
-release = '3.8.0'
+release = '3.9.0'
# The short X.Y version.
version = '.'.join(release.split('.')[:2])
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/docs/high/dataset.rst
new/h5py-3.9.0/docs/high/dataset.rst
--- old/h5py-3.8.0/docs/high/dataset.rst 2023-01-04 18:26:33.000000000
+0100
+++ new/h5py-3.9.0/docs/high/dataset.rst 2023-04-26 17:08:31.000000000
+0200
@@ -503,12 +503,8 @@
>>> out.dtype
dtype('int16')
- .. versionchanged:: 3.0
- Allowed reading through the wrapper object. In earlier versions,
- :meth:`astype` had to be used as a context manager:
-
- >>> with dset.astype('int16'):
- ... out = dset[:]
+ .. versionchanged:: 3.9
+ :meth:`astype` can no longer be used as a context manager.
.. method:: asstr(encoding=None, errors='strict')
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/docs/high/file.rst
new/h5py-3.9.0/docs/high/file.rst
--- old/h5py-3.8.0/docs/high/file.rst 2023-01-04 18:26:33.000000000 +0100
+++ new/h5py-3.9.0/docs/high/file.rst 2023-02-17 11:43:33.000000000 +0100
@@ -115,6 +115,13 @@
The argument values must be ``bytes`` objects. All three arguments are
required to activate AWS authentication.
+ .. note::
+ Pre-built h5py packages on PyPI do not include this S3 support. If
+ you want this feature, you could use packages from conda-forge, or
+ :ref:`build h5py from source <source_install>` against an HDF5 build
+ with S3 support. Alternatively, use the :ref:`file-like object
+ <file_fileobj>` support with a package like s3fs.
+
.. _file_fileobj:
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/docs/whatsnew/3.9.rst
new/h5py-3.9.0/docs/whatsnew/3.9.rst
--- old/h5py-3.8.0/docs/whatsnew/3.9.rst 1970-01-01 01:00:00.000000000
+0100
+++ new/h5py-3.9.0/docs/whatsnew/3.9.rst 2023-06-20 10:56:46.000000000
+0200
@@ -0,0 +1,56 @@
+What's new in h5py 3.9
+======================
+
+This version of h5py requires Python 3.8 or above.
+
+New features
+------------
+
+* New ``out`` argument to :meth:`~h5py.h5d.DatasetID.read_direct_chunk` to
allow passing
+ the output buffer (:pr:`2232`).
+* The objects from :meth:`.Dataset.asstr` and :meth:`.Dataset.astype` now
+ implement the ``__array__()`` method (:pr:`2269`).
+ This speeds up access for functions that support it, such as
``np.asarray()``.
+* Validate key types when creating groups and attributes, giving better error
+ messages when invalid types are used (:pr:`2266`).
+
+Deprecations & removals
+-----------------------
+
+* Using :meth:`.Dataset.astype` as a context manager has been removed, after
+ being deprecated in h5py 3.6. Read data by slicing the returned object
instead:
+ ``dset.astype('f4')[:]``.
+
+Exposing HDF5 functions
+-----------------------
+
+* ``H5Pget_elink_acc_flags`` & ``H5Pset_elink_acc_flags`` as
+ :meth:`h5py.h5p.PropLAID.get_elink_acc_flags` &
:meth:`h5py.h5p.PropLAID.set_elink_acc_flags`:
+ access the external link file access traversal flags in a link access
property
+ list (:pr:`2244`).
+* ``H5Zregister`` as :func:`h5py.h5z.register_filter`: register an HDF5 filter
+ (:pr:`2229`).
+
+Bug fixes
+---------
+
+* ``Group.__contains__`` and ``Group.get`` now use the default link access
+ property list systematically (:pr:`2244`).
+* Removed various calls to the deprecated ``numpy.product`` function
(:pr:`2242`
+ & :pr:`2273`).
+* Fix the IPython tab-completion integration in IPython 8.12 (:pr:2256`).
+* Replacing attributes with :meth:`.AttributeManager.create` now deletes the
old
+ attributes before creating the new one, rather than using a temporary name
+ and renaming the new attribute (:pr:`2274`). This should avoid some confusing
+ bugs affecting attributes. However, failures creating an attribute are less
+ likely to leave an existing attribute of the same name in place. To change an
+ attribute value without changing its shape or dtype, use
+ :meth:`~.AttributeManager.modify` instead.
+
+Building h5py
+-------------
+
+* When building with :ref:`parallel` support, the version of mpi4py used on
+ various Python versions is increased to 3.1.1, fixing building with a newer
+ setuptools (:pr:`2225`).
+* Some fixes towards compatibility with the upcoming Cython 3 (:pr:`2247`).
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/docs/whatsnew/index.rst
new/h5py-3.9.0/docs/whatsnew/index.rst
--- old/h5py-3.8.0/docs/whatsnew/index.rst 2023-01-23 11:18:46.000000000
+0100
+++ new/h5py-3.9.0/docs/whatsnew/index.rst 2023-06-20 10:56:46.000000000
+0200
@@ -8,6 +8,7 @@
.. toctree::
+ 3.9
3.8
3.7
3.6
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py/_errors.pyx
new/h5py-3.9.0/h5py/_errors.pyx
--- old/h5py-3.8.0/h5py/_errors.pyx 2023-01-04 18:26:33.000000000 +0100
+++ new/h5py-3.9.0/h5py/_errors.pyx 2023-04-26 17:08:31.000000000 +0200
@@ -94,7 +94,7 @@
H5E_error_t err
int n
-cdef herr_t walk_cb(unsigned int n, const H5E_error_t *desc, void *e) nogil:
+cdef herr_t walk_cb(unsigned int n, const H5E_error_t *desc, void *e) nogil
noexcept:
cdef err_data_t *ee = <err_data_t*>e
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py/_hl/attrs.py
new/h5py-3.9.0/h5py/_hl/attrs.py
--- old/h5py-3.8.0/h5py/_hl/attrs.py 2023-01-04 18:26:33.000000000 +0100
+++ new/h5py-3.9.0/h5py/_hl/attrs.py 2023-06-16 12:09:33.000000000 +0200
@@ -122,6 +122,7 @@
Data type of the attribute. Overrides data.dtype if both
are given.
"""
+ name = self._e(name)
with phil:
# First, make sure we have a NumPy array. We leave the data type
@@ -166,7 +167,7 @@
# is compatible, and reshape if needed.
else:
- if shape is not None and numpy.product(shape,
dtype=numpy.ulonglong) != numpy.product(data.shape, dtype=numpy.ulonglong):
+ if shape is not None and product(shape) != product(data.shape):
raise ValueError("Shape of new attribute conflicts with
shape of data")
if shape != data.shape:
@@ -189,35 +190,24 @@
else:
space = h5s.create_simple(shape)
- # This mess exists because you can't overwrite attributes in HDF5.
- # So we write to a temporary attribute first, and then rename.
- # see issue 1385
- # if track_order is enabled new attributes (which exceed the
- # max_compact range, 8 is default) cannot be created as temporary
- # attributes with subsequent rename, doing that would trigger
- # the error discussed in the above issue
- attr_exists = False
- if h5a.exists(self._id, self._e(name)):
- attr_exists = True
- tempname = uuid.uuid4().hex
- else:
- tempname = name
+ # For a long time, h5py would create attributes with a random name
+ # and then rename them, imitating how you can atomically replace
+ # a file in a filesystem. But HDF5 does not offer atomic
replacement
+ # (you have to delete the existing attribute first), and renaming
+ # exposes some bugs - see https://github.com/h5py/h5py/issues/1385
+ # So we've gone back to the simpler delete & recreate model.
+ if h5a.exists(self._id, name):
+ h5a.delete(self._id, name)
- attr = h5a.create(self._id, self._e(tempname), htype, space)
+ attr = h5a.create(self._id, name, htype, space)
try:
if not isinstance(data, Empty):
attr.write(data, mtype=htype2)
- if attr_exists:
- # Rename temp attribute to proper name
- # No atomic rename in HDF5 :(
- h5a.delete(self._id, self._e(name))
- h5a.rename(self._id, self._e(tempname), self._e(name))
except:
attr.close()
- h5a.delete(self._id, self._e(tempname))
+ h5a.delete(self._id, name)
raise
- finally:
- attr.close()
+ attr.close()
def modify(self, name, value):
""" Change the value of an attribute while preserving its type.
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py/_hl/base.py
new/h5py-3.9.0/h5py/_hl/base.py
--- old/h5py-3.8.0/h5py/_hl/base.py 2022-12-21 17:13:47.000000000 +0100
+++ new/h5py-3.9.0/h5py/_hl/base.py 2023-06-07 13:39:16.000000000 +0200
@@ -195,13 +195,15 @@
if isinstance(name, bytes):
coding = h5t.CSET_ASCII
- else:
+ elif isinstance(name, str):
try:
name = name.encode('ascii')
coding = h5t.CSET_ASCII
except UnicodeEncodeError:
name = name.encode('utf8')
coding = h5t.CSET_UTF8
+ else:
+ raise TypeError(f"A name should be string or bytes, not
{type(name)}")
if lcpl:
return name, get_lcpl(coding)
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py/_hl/dataset.py
new/h5py-3.9.0/h5py/_hl/dataset.py
--- old/h5py-3.8.0/h5py/_hl/dataset.py 2023-01-04 18:26:33.000000000 +0100
+++ new/h5py-3.9.0/h5py/_hl/dataset.py 2023-06-16 12:09:33.000000000 +0200
@@ -13,15 +13,14 @@
import posixpath as pp
import sys
-from warnings import warn
-
-from threading import local
import numpy
from .. import h5, h5s, h5t, h5r, h5d, h5p, h5fd, h5ds, _selector
-from ..h5py_warnings import H5pyDeprecationWarning
-from .base import HLObject, phil, with_phil, Empty, cached_property,
find_item_type
+from .base import (
+ array_for_new_object, cached_property, Empty, find_item_type, HLObject,
+ phil, product, with_phil,
+)
from . import filters
from . import selections as sel
from . import selections2 as sel2
@@ -44,8 +43,7 @@
# Convert data to a C-contiguous ndarray
if data is not None and not isinstance(data, Empty):
- from . import base
- data = base.array_for_new_object(data, specified_dtype=dtype)
+ data = array_for_new_object(data, specified_dtype=dtype)
# Validate shape
if shape is None:
@@ -56,7 +54,7 @@
shape = data.shape
else:
shape = (shape,) if isinstance(shape, int) else tuple(shape)
- if data is not None and (numpy.product(shape, dtype=numpy.ulonglong)
!= numpy.product(data.shape, dtype=numpy.ulonglong)):
+ if data is not None and (product(shape) != product(data.shape)):
raise ValueError("Shape tuple is incompatible with data")
if isinstance(maxshape, int):
@@ -208,20 +206,6 @@
def __getitem__(self, args):
return self._dset.__getitem__(args, new_dtype=self._dtype)
- def __enter__(self):
- # pylint: disable=protected-access
- warn(
- "Using astype() as a context manager is deprecated. "
- "Slice the returned object instead, like:
ds.astype(np.int32)[:10]",
- category=H5pyDeprecationWarning, stacklevel=2,
- )
- self._dset._local.astype = self._dtype
- return self
-
- def __exit__(self, *args):
- # pylint: disable=protected-access
- self._dset._local.astype = None
-
def __len__(self):
""" Get the length of the underlying dataset
@@ -229,6 +213,12 @@
"""
return len(self._dset)
+ def __array__(self, dtype=None):
+ data = self[:]
+ if dtype is not None:
+ data = data.astype(dtype)
+ return data
+
class AsStrWrapper:
"""Wrapper to decode strings on reading the dataset"""
@@ -261,6 +251,11 @@
"""
return len(self._dset)
+ def __array__(self):
+ return numpy.array([
+ b.decode(self.encoding, self.errors) for b in self._dset
+ ], dtype=object).reshape(self._dset.shape)
+
class FieldsWrapper:
"""Wrapper to extract named fields from a dataset with a struct dtype"""
@@ -494,7 +489,7 @@
if self._is_empty:
size = None
else:
- size = numpy.prod(self.shape, dtype=numpy.intp)
+ size = product(self.shape)
# If the file is read-only, cache the size to speed-up future uses.
# This cache is invalidated by .refresh() when using SWMR.
@@ -651,8 +646,6 @@
self._filters = filters.get_filters(self._dcpl)
self._readonly = readonly
self._cache_props = {}
- self._local = local()
- self._local.astype = None
def resize(self, size, axis=None):
""" Resize the dataset, or the specified axis.
@@ -760,9 +753,6 @@
"""
args = args if isinstance(args, tuple) else (args,)
- if new_dtype is None:
- new_dtype = getattr(self._local, 'astype', None)
-
if self._fast_read_ok and (new_dtype is None):
try:
return self._fast_reader.read(args)
@@ -885,7 +875,8 @@
if val.ndim > 1:
tmp = numpy.empty(shape=val.shape[:-1], dtype=object)
tmp.ravel()[:] = [i for i in val.reshape(
- (numpy.product(val.shape[:-1], dtype=numpy.ulonglong),
val.shape[-1]))]
+ (product(val.shape[:-1]), val.shape[-1])
+ )]
else:
tmp = numpy.array([None], dtype=object)
tmp[0] = val
@@ -994,8 +985,7 @@
if mshape == () and selection.array_shape != ():
if self.dtype.subdtype is not None:
raise TypeError("Scalar broadcasting is not supported for
array dtypes")
- if self.chunks and (numpy.prod(self.chunks, dtype=numpy.float64) >=
- numpy.prod(selection.array_shape,
dtype=numpy.float64)):
+ if self.chunks and (product(self.chunks) >=
product(selection.array_shape)):
val2 = numpy.empty(selection.array_shape, dtype=val.dtype)
else:
val2 = numpy.empty(selection.array_shape[-1], dtype=val.dtype)
@@ -1067,7 +1057,7 @@
arr = numpy.zeros(self.shape, dtype=self.dtype if dtype is None else
dtype)
# Special case for (0,)*-shape datasets
- if numpy.product(self.shape, dtype=numpy.ulonglong) == 0:
+ if self.size == 0:
return arr
self.read_direct(arr)
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py/_hl/filters.py
new/h5py-3.9.0/h5py/_hl/filters.py
--- old/h5py-3.8.0/h5py/_hl/filters.py 2022-12-21 17:13:47.000000000 +0100
+++ new/h5py-3.9.0/h5py/_hl/filters.py 2023-04-26 17:08:31.000000000 +0200
@@ -41,6 +41,7 @@
import operator
import numpy as np
+from .base import product
from .compat import filename_encode
from .. import h5z, h5p, h5d, h5f
@@ -358,7 +359,7 @@
# Determine the optimal chunk size in bytes using a PyTables expression.
# This is kept as a float.
- dset_size = np.product(chunks)*typesize
+ dset_size = product(chunks)*typesize
target_size = CHUNK_BASE * (2**np.log10(dset_size/(1024.*1024)))
if target_size > CHUNK_MAX:
@@ -373,14 +374,14 @@
# 1b. We're within 50% of the target chunk size, AND
# 2. The chunk is smaller than the maximum chunk size
- chunk_bytes = np.product(chunks)*typesize
+ chunk_bytes = product(chunks)*typesize
if (chunk_bytes < target_size or \
abs(chunk_bytes-target_size)/target_size < 0.5) and \
chunk_bytes < CHUNK_MAX:
break
- if np.product(chunks) == 1:
+ if product(chunks) == 1:
break # Element size larger than CHUNK_MAX
chunks[idx%ndims] = np.ceil(chunks[idx%ndims] / 2.0)
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py/_hl/group.py
new/h5py-3.9.0/h5py/_hl/group.py
--- old/h5py-3.8.0/h5py/_hl/group.py 2023-01-04 18:26:33.000000000 +0100
+++ new/h5py-3.9.0/h5py/_hl/group.py 2023-04-26 17:08:31.000000000 +0200
@@ -405,7 +405,7 @@
return default
elif getclass and not getlink:
- typecode = h5o.get_info(self.id, self._e(name)).type
+ typecode = h5o.get_info(self.id, self._e(name),
lapl=self._lapl).type
try:
return {h5o.TYPE_GROUP: Group,
@@ -415,18 +415,18 @@
raise TypeError("Unknown object type")
elif getlink:
- typecode = self.id.links.get_info(self._e(name)).type
+ typecode = self.id.links.get_info(self._e(name),
lapl=self._lapl).type
if typecode == h5l.TYPE_SOFT:
if getclass:
return SoftLink
- linkbytes = self.id.links.get_val(self._e(name))
+ linkbytes = self.id.links.get_val(self._e(name),
lapl=self._lapl)
return SoftLink(self._d(linkbytes))
elif typecode == h5l.TYPE_EXTERNAL:
if getclass:
return ExternalLink
- filebytes, linkbytes = self.id.links.get_val(self._e(name))
+ filebytes, linkbytes =
self.id.links.get_val(self._e(name), lapl=self._lapl)
return ExternalLink(
filename_decode(filebytes), self._d(linkbytes)
)
@@ -508,6 +508,10 @@
@with_phil
def __contains__(self, name):
""" Test if a member name exists """
+ if hasattr(h5g, "_path_valid"):
+ if not self.id:
+ return False
+ return h5g._path_valid(self.id, self._e(name), self._lapl)
return self._e(name) in self.id
def copy(self, source, dest, name=None,
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py/_hl/selections.py
new/h5py-3.9.0/h5py/_hl/selections.py
--- old/h5py-3.8.0/h5py/_hl/selections.py 2022-12-21 17:13:47.000000000
+0100
+++ new/h5py-3.9.0/h5py/_hl/selections.py 2023-04-26 17:08:31.000000000
+0200
@@ -427,7 +427,7 @@
shape = tuple(get_n_axis(sid, x) for x in range(rank))
- if np.product(shape) != N:
+ if product(shape) != N:
# This means multiple hyperslab selections are in effect,
# so we fall back to a 1D shape
return (N,)
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py/api_functions.txt
new/h5py-3.9.0/h5py/api_functions.txt
--- old/h5py-3.8.0/h5py/api_functions.txt 2023-01-10 18:06:37.000000000
+0100
+++ new/h5py-3.9.0/h5py/api_functions.txt 2023-04-26 17:08:31.000000000
+0200
@@ -391,6 +391,8 @@
ssize_t H5Pget_elink_prefix(hid_t plist_id, char *prefix, size_t size)
hid_t H5Pget_elink_fapl(hid_t lapl_id)
herr_t H5Pset_elink_fapl(hid_t lapl_id, hid_t fapl_id)
+ herr_t H5Pget_elink_acc_flags(hid_t lapl_id, unsigned int *flags)
+ herr_t H5Pset_elink_acc_flags(hid_t lapl_id, unsigned int flags)
herr_t H5Pset_create_intermediate_group(hid_t plist_id, unsigned
crt_intmd)
herr_t H5Pget_create_intermediate_group(hid_t plist_id, unsigned
*crt_intmd)
@@ -586,6 +588,7 @@
htri_t H5Zfilter_avail(H5Z_filter_t id_)
herr_t H5Zget_filter_info(H5Z_filter_t filter_, unsigned int
*filter_config_flags)
+ herr_t H5Zregister(const void *cls)
herr_t H5Zunregister(H5Z_filter_t id_)
hdf5_hl:
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py/api_types_hdf5.pxd
new/h5py-3.9.0/h5py/api_types_hdf5.pxd
--- old/h5py-3.8.0/h5py/api_types_hdf5.pxd 2023-01-10 18:06:37.000000000
+0100
+++ new/h5py-3.9.0/h5py/api_types_hdf5.pxd 2023-04-26 17:08:31.000000000
+0200
@@ -776,6 +776,7 @@
ctypedef int H5Z_filter_t
+ int H5Z_CLASS_T_VERS
int H5Z_FILTER_ERROR
int H5Z_FILTER_NONE
int H5Z_FILTER_ALL
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py/h5d.pyx new/h5py-3.9.0/h5py/h5d.pyx
--- old/h5py-3.8.0/h5py/h5d.pyx 2023-01-10 18:06:37.000000000 +0100
+++ new/h5py-3.9.0/h5py/h5d.pyx 2023-04-26 17:08:31.000000000 +0200
@@ -16,6 +16,7 @@
# Compile-time imports
from collections import namedtuple
+cimport cython
from ._objects cimport pdefault
from numpy cimport ndarray, import_array, PyArray_DATA
from .utils cimport check_numpy_read, check_numpy_write, \
@@ -26,9 +27,11 @@
from ._proxy cimport dset_rw
from ._objects import phil, with_phil
-from cpython cimport PyObject_GetBuffer, \
- PyBUF_ANY_CONTIGUOUS, \
- PyBuffer_Release
+from cpython cimport PyBUF_ANY_CONTIGUOUS, \
+ PyBuffer_Release, \
+ PyBytes_AsString, \
+ PyBytes_FromStringAndSize, \
+ PyObject_GetBuffer
# Initialization
@@ -516,16 +519,18 @@
IF HDF5_VERSION >= (1, 10, 2):
- def read_direct_chunk(self, offsets, PropID dxpl=None):
- """ (offsets, PropID dxpl=None)
+ @cython.boundscheck(False)
+ @cython.wraparound(False)
+ def read_direct_chunk(self, offsets, PropID dxpl=None, unsigned
char[::1] out=None):
+ """ (offsets, PropID dxpl=None, out=None)
Reads data to a bytes array directly from a chunk at position
specified by the `offsets` argument and bypasses any filters HDF5
would normally apply to the written data. However, the written data
may be compressed or not.
- Returns a tuple containing the `filter_mask` and the bytes data
- which are the raw data storing this chuck.
+ Returns a tuple containing the `filter_mask` and the raw data
+ storing this chunk as bytes if `out` is None, else as a memoryview.
`filter_mask` is a bit field of up to 32 values. It records which
filters have been applied to this chunk, of the filter pipeline
@@ -534,47 +539,59 @@
compute the raw data. So the default value of `0` means that all
defined filters have been applied to the raw data.
+ If the `out` argument is not None, it must be a writeable
+ contiguous 1D array-like of bytes (e.g., `bytearray` or
+ `numpy.ndarray`) and large enough to contain the whole chunk.
+
Feature requires: 1.10.2 HDF5
"""
-
cdef hid_t dset_id
cdef hid_t dxpl_id
- cdef hid_t space_id = 0
+ cdef hid_t space_id
cdef hsize_t *offset = NULL
- cdef size_t data_size
cdef int rank
cdef uint32_t filters
- cdef hsize_t read_chunk_nbytes
- cdef char *data = NULL
- cdef bytes ret
+ cdef hsize_t chunk_bytes, out_bytes
+ cdef int nb_offsets = len(offsets)
+ cdef void * chunk_buffer
dset_id = self.id
dxpl_id = pdefault(dxpl)
- space_id = H5Dget_space(self.id)
+ space_id = H5Dget_space(dset_id)
rank = H5Sget_simple_extent_ndims(space_id)
+ H5Sclose(space_id)
- if len(offsets) != rank:
- raise TypeError("offset length (%d) must match dataset rank
(%d)" % (len(offsets), rank))
+ if nb_offsets != rank:
+ raise TypeError(
+ f"offsets length ({nb_offsets}) must match dataset rank
({rank})"
+ )
+ offset = <hsize_t*>emalloc(sizeof(hsize_t)*rank)
try:
- offset = <hsize_t*>emalloc(sizeof(hsize_t)*rank)
convert_tuple(offsets, offset, rank)
- H5Dget_chunk_storage_size(dset_id, offset, &read_chunk_nbytes)
- data = <char *>emalloc(read_chunk_nbytes)
+ H5Dget_chunk_storage_size(dset_id, offset, &chunk_bytes)
+
+ if out is None:
+ retval = PyBytes_FromStringAndSize(NULL, chunk_bytes)
+ chunk_buffer = PyBytes_AsString(retval)
+ else:
+ out_bytes = out.shape[0] # Fast way to get out length
+ if out_bytes < chunk_bytes:
+ raise ValueError(
+ f"out buffer is only {out_bytes} bytes,
{chunk_bytes} bytes required"
+ )
+ retval = memoryview(out[:chunk_bytes])
+ chunk_buffer = &out[0]
IF HDF5_VERSION >= (1, 10, 3):
- H5Dread_chunk(dset_id, dxpl_id, offset, &filters, data)
+ H5Dread_chunk(dset_id, dxpl_id, offset, &filters,
chunk_buffer)
ELSE:
- H5DOread_chunk(dset_id, dxpl_id, offset, &filters, data)
- ret = data[:read_chunk_nbytes]
+ H5DOread_chunk(dset_id, dxpl_id, offset, &filters,
chunk_buffer)
finally:
efree(offset)
- if data:
- efree(data)
- if space_id:
- H5Sclose(space_id)
- return filters, ret
+ return filters, retval
+
IF HDF5_VERSION >= (1, 10, 5):
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py/h5p.pyx new/h5py-3.9.0/h5py/h5p.pyx
--- old/h5py-3.8.0/h5py/h5p.pyx 2023-01-04 18:26:33.000000000 +0100
+++ new/h5py-3.9.0/h5py/h5p.pyx 2023-04-26 17:08:31.000000000 +0200
@@ -1644,6 +1644,27 @@
H5Idec_ref(fid)
return propwrap(fid)
+
+ @with_phil
+ def set_elink_acc_flags(self, unsigned int flags):
+ """ (UNIT flags)
+
+ Sets the external link traversal file access flag in a link access
property list.
+ """
+ H5Pset_elink_acc_flags(self.id, flags)
+
+
+ @with_phil
+ def get_elink_acc_flags(self):
+ """() => UINT
+
+ Retrieves the external link traversal file access flag from the
specified link access property list.
+ """
+ cdef unsigned int flags
+ H5Pget_elink_acc_flags(self.id, &flags)
+ return flags
+
+
# Datatype creation
cdef class PropTCID(PropOCID):
""" Datatype creation property list
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py/h5z.pyx new/h5py-3.9.0/h5py/h5z.pyx
--- old/h5py-3.8.0/h5py/h5z.pyx 2021-05-09 13:55:36.000000000 +0200
+++ new/h5py-3.9.0/h5py/h5z.pyx 2023-04-26 17:08:31.000000000 +0200
@@ -16,6 +16,8 @@
# === Public constants and data structures ====================================
+CLASS_T_VERS = H5Z_CLASS_T_VERS
+
FILTER_LZF = H5PY_FILTER_LZF
FILTER_ERROR = H5Z_FILTER_ERROR
@@ -97,6 +99,27 @@
@with_phil
+def register_filter(Py_ssize_t cls_pointer_address):
+ '''(INT cls_pointer_address) => BOOL
+
+ Register a new filter from the memory address of a buffer containing a
+ ``H5Z_class1_t`` or ``H5Z_class2_t`` data structure describing the filter.
+
+ `cls_pointer_address` can be retrieved from a HDF5 filter plugin dynamic
+ library::
+
+ import ctypes
+
+ filter_clib = ctypes.CDLL("/path/to/my_hdf5_filter_plugin.so")
+ filter_clib.H5PLget_plugin_info.restype = ctypes.c_void_p
+
+ h5py.h5z.register_filter(filter_clib.H5PLget_plugin_info())
+
+ '''
+ return <int>H5Zregister(<const void *>cls_pointer_address) >= 0
+
+
+@with_phil
def unregister_filter(int filter_code):
'''(INT filter_code) => BOOL
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py/ipy_completer.py
new/h5py-3.9.0/h5py/ipy_completer.py
--- old/h5py-3.8.0/h5py/ipy_completer.py 2022-12-21 17:13:47.000000000
+0100
+++ new/h5py-3.9.0/h5py/ipy_completer.py 2023-05-16 18:24:09.000000000
+0200
@@ -128,7 +128,12 @@
""" Completer function to be loaded into IPython """
base = re_object_match.split(event.line)[1]
- if not isinstance(self._ofind(base).get('obj'), (AttributeManager,
HLObject)):
+ try:
+ obj = self._ofind(base).obj
+ except AttributeError:
+ obj = self._ofind(base).get('obj')
+
+ if not isinstance(obj, (AttributeManager, HLObject)):
raise TryNext
try:
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py/tests/test_attribute_create.py
new/h5py-3.9.0/h5py/tests/test_attribute_create.py
--- old/h5py-3.8.0/h5py/tests/test_attribute_create.py 2021-05-09
13:55:36.000000000 +0200
+++ new/h5py-3.9.0/h5py/tests/test_attribute_create.py 2023-06-07
13:39:16.000000000 +0200
@@ -89,3 +89,7 @@
dt = np.dtype('()i')
with self.assertRaises(ValueError):
self.f.attrs.create('x', data=array, shape=(5,), dtype=dt)
+
+ def test_key_type(self):
+ with self.assertRaises(TypeError):
+ self.f.attrs.create(1, data=('a', 'b'))
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py/tests/test_dataset.py
new/h5py-3.9.0/h5py/tests/test_dataset.py
--- old/h5py-3.8.0/h5py/tests/test_dataset.py 2023-01-10 18:06:37.000000000
+0100
+++ new/h5py-3.9.0/h5py/tests/test_dataset.py 2023-06-07 13:39:16.000000000
+0200
@@ -27,7 +27,7 @@
from .common import ut, TestCase
from .data_files import get_data_file_path
from h5py import File, Group, Dataset
-from h5py._hl.base import is_empty_dataspace
+from h5py._hl.base import is_empty_dataspace, product
from h5py import h5f, h5t
from h5py.h5py_warnings import H5pyDeprecationWarning
from h5py import version
@@ -231,7 +231,7 @@
((5, 7, 9), (6,), np.s_[2, :6, 3], np.s_[:]),
])
def test_read_direct(self, writable_file, source_shape, dest_shape,
source_sel, dest_sel):
- source_values = np.arange(np.product(source_shape),
dtype="int64").reshape(source_shape)
+ source_values = np.arange(product(source_shape),
dtype="int64").reshape(source_shape)
dset = writable_file.create_dataset("dset", source_shape,
data=source_values)
arr = np.full(dest_shape, -1, dtype="int64")
expected = arr.copy()
@@ -280,7 +280,7 @@
])
def test_write_direct(self, writable_file, source_shape, dest_shape,
source_sel, dest_sel):
dset = writable_file.create_dataset('dset', dest_shape, dtype='int32',
fillvalue=-1)
- arr = np.arange(np.product(source_shape)).reshape(source_shape)
+ arr = np.arange(product(source_shape)).reshape(source_shape)
expected = np.full(dest_shape, -1, dtype='int32')
expected[dest_sel] = arr[source_sel]
dset.write_direct(arr, source_sel, dest_sel)
@@ -1257,12 +1257,15 @@
# len of ds
self.assertEqual(10, len(ds.asstr()))
-
# Array output
np.testing.assert_array_equal(
ds.asstr()[:1], np.array([data], dtype=object)
)
+ np.testing.assert_array_equal(
+ np.asarray(ds.asstr())[:1], np.array([data], dtype=object)
+ )
+
def test_asstr_fixed(self):
dt = h5py.string_dtype(length=5)
ds = self.f.create_dataset('x', (10,), dtype=dt)
@@ -1569,21 +1572,6 @@
class TestAstype(BaseDataset):
""".astype() wrapper & context manager
"""
- def test_astype_ctx(self):
- dset = self.f.create_dataset('x', (100,), dtype='i2')
- dset[...] = np.arange(100)
-
- with warnings.catch_warnings(record=True) as warn_rec:
- warnings.simplefilter("always")
-
- with dset.astype('f8'):
- self.assertArrayEqual(dset[...], np.arange(100, dtype='f8'))
-
- with dset.astype('f4') as f4ds:
- self.assertArrayEqual(f4ds[...], np.arange(100, dtype='f4'))
-
- assert [w.category for w in warn_rec] == [H5pyDeprecationWarning] * 2
-
def test_astype_wrapper(self):
dset = self.f.create_dataset('x', (100,), dtype='i2')
dset[...] = np.arange(100)
@@ -1596,6 +1584,12 @@
dset[...] = np.arange(100)
self.assertEqual(100, len(dset.astype('f4')))
+ def test_astype_wrapper_asarray(self):
+ dset = self.f.create_dataset('x', (100,), dtype='i2')
+ dset[...] = np.arange(100)
+ arr = np.asarray(dset.astype('f4'), dtype='i2')
+ self.assertArrayEqual(arr, np.arange(100, dtype='i2'))
+
class TestScalarCompound(BaseDataset):
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py/tests/test_filters.py
new/h5py-3.9.0/h5py/tests/test_filters.py
--- old/h5py-3.8.0/h5py/tests/test_filters.py 2022-12-21 17:13:47.000000000
+0100
+++ new/h5py-3.9.0/h5py/tests/test_filters.py 2023-04-26 17:08:31.000000000
+0200
@@ -14,9 +14,8 @@
import os
import numpy as np
import h5py
-import pytest
-from .common import ut, TestCase, insubprocess
+from .common import ut, TestCase
class TestFilters(TestCase):
@@ -87,14 +86,6 @@
assert gzip8 != h5py.filters.Gzip(level=7)
-@pytest.mark.mpi_skip
-@insubprocess
-def test_unregister_filter(request):
- if h5py.h5z.filter_avail(h5py.h5z.FILTER_LZF):
- res = h5py.h5z.unregister_filter(h5py.h5z.FILTER_LZF)
- assert res
-
-
@ut.skipIf(not os.getenv('H5PY_TEST_CHECK_FILTERS'), "H5PY_TEST_CHECK_FILTERS
not set")
def test_filters_available():
assert 'gzip' in h5py.filters.decode
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py/tests/test_group.py
new/h5py-3.9.0/h5py/tests/test_group.py
--- old/h5py-3.8.0/h5py/tests/test_group.py 2022-12-21 17:13:47.000000000
+0100
+++ new/h5py-3.9.0/h5py/tests/test_group.py 2023-06-07 13:39:16.000000000
+0200
@@ -94,6 +94,11 @@
self.assertEqual(group.name, name)
self.assertEqual(group.id.links.get_info(name.encode('utf8')).cset,
h5t.CSET_ASCII)
+ def test_type(self):
+ """ Names should be strings or bytes """
+ with self.assertRaises(TypeError):
+ self.f.create_group(1.)
+
def test_appropriate_low_level_id(self):
" Binding a group to a non-group identifier fails with ValueError "
dset = self.f.create_dataset('foo', [1])
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py/tests/test_h5d_direct_chunk.py
new/h5py-3.9.0/h5py/tests/test_h5d_direct_chunk.py
--- old/h5py-3.8.0/h5py/tests/test_h5d_direct_chunk.py 2022-12-21
17:13:47.000000000 +0100
+++ new/h5py-3.9.0/h5py/tests/test_h5d_direct_chunk.py 2023-04-26
17:08:31.000000000 +0200
@@ -1,6 +1,7 @@
import h5py
import numpy
import numpy.testing
+import pytest
from .common import ut, TestCase
@@ -113,3 +114,81 @@
with h5py.File(filename, "r") as filehandle:
dataset = filehandle["created"][...]
numpy.testing.assert_array_equal(dataset, frame)
+
+
+@pytest.mark.skipif(
+ h5py.version.hdf5_version_tuple < (1, 10, 2),
+ reason="Direct chunk read requires HDF5 >= 1.10.2"
+)
+class TestReadDirectChunkToOut:
+
+ def test_uncompressed_data(self, writable_file):
+ ref_data = numpy.arange(16).reshape(4, 4)
+ dataset = writable_file.create_dataset(
+ "uncompressed", data=ref_data, chunks=ref_data.shape)
+
+ out = bytearray(ref_data.nbytes)
+ filter_mask, chunk = dataset.id.read_direct_chunk((0, 0), out=out)
+
+ assert numpy.array_equal(
+ numpy.frombuffer(out,
dtype=ref_data.dtype).reshape(ref_data.shape),
+ ref_data,
+ )
+ assert filter_mask == 0
+ assert len(chunk) == ref_data.nbytes
+
+ @pytest.mark.skipif(
+ h5py.version.hdf5_version_tuple < (1, 10, 5),
+ reason="chunk info requires HDF5 >= 1.10.5",
+ )
+ @pytest.mark.skipif(
+ 'gzip' not in h5py.filters.encode,
+ reason="DEFLATE is not installed",
+ )
+ def test_compressed_data(self, writable_file):
+ ref_data = numpy.arange(16).reshape(4, 4)
+ dataset = writable_file.create_dataset(
+ "gzip",
+ data=ref_data,
+ chunks=ref_data.shape,
+ compression="gzip",
+ compression_opts=9,
+ )
+ chunk_info = dataset.id.get_chunk_info(0)
+
+ out = bytearray(chunk_info.size)
+ filter_mask, chunk = dataset.id.read_direct_chunk(
+ chunk_info.chunk_offset,
+ out=out,
+ )
+ assert filter_mask == chunk_info.filter_mask
+ assert len(chunk) == chunk_info.size
+ assert out == dataset.id.read_direct_chunk(chunk_info.chunk_offset)[1]
+
+ def test_fail_buffer_too_small(self, writable_file):
+ ref_data = numpy.arange(16).reshape(4, 4)
+ dataset = writable_file.create_dataset(
+ "uncompressed", data=ref_data, chunks=ref_data.shape)
+
+ out = bytearray(ref_data.nbytes // 2)
+ with pytest.raises(ValueError):
+ dataset.id.read_direct_chunk((0, 0), out=out)
+
+ def test_fail_buffer_readonly(self, writable_file):
+ ref_data = numpy.arange(16).reshape(4, 4)
+ dataset = writable_file.create_dataset(
+ "uncompressed", data=ref_data, chunks=ref_data.shape)
+
+ out = bytes(ref_data.nbytes)
+ with pytest.raises(BufferError):
+ dataset.id.read_direct_chunk((0, 0), out=out)
+
+ def test_fail_buffer_not_contiguous(self, writable_file):
+ ref_data = numpy.arange(16).reshape(4, 4)
+ dataset = writable_file.create_dataset(
+ "uncompressed", data=ref_data, chunks=ref_data.shape)
+
+ array = numpy.empty(ref_data.shape + (2,), dtype=ref_data.dtype)
+ out = array[:, :, ::2] # Array is not contiguous
+ with pytest.raises(ValueError):
+ dataset.id.read_direct_chunk((0, 0), out=out)
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py/tests/test_h5z.py
new/h5py-3.9.0/h5py/tests/test_h5z.py
--- old/h5py-3.8.0/h5py/tests/test_h5z.py 1970-01-01 01:00:00.000000000
+0100
+++ new/h5py-3.9.0/h5py/tests/test_h5z.py 2023-04-26 17:08:31.000000000
+0200
@@ -0,0 +1,85 @@
+from ctypes import (
+ addressof,
+ c_char_p,
+ c_int,
+ c_long,
+ c_uint,
+ c_void_p,
+ CFUNCTYPE,
+ POINTER,
+ Structure,
+)
+import pytest
+import h5py
+from h5py import h5z
+
+from .common import insubprocess
+
+
+# Type of filter callback function of H5Z_class2_t
+H5ZFuncT = CFUNCTYPE(
+ c_long, # restype
+ # argtypes
+ c_uint, # flags
+ c_long, # cd_nelemts
+ POINTER(c_uint), # cd_values
+ c_long, # nbytes
+ POINTER(c_long), # buf_size
+ POINTER(c_void_p), # buf
+)
+
+
+class H5ZClass2T(Structure):
+ """H5Z_class2_t structure defining a filter"""
+
+ _fields_ = [
+ ("version", c_int),
+ ("id_", c_int),
+ ("encoder_present", c_uint),
+ ("decoder_present", c_uint),
+ ("name", c_char_p),
+ ("can_apply", c_void_p),
+ ("set_local", c_void_p),
+ ("filter_", H5ZFuncT),
+ ]
+
+
+def test_register_filter():
+ filter_id = 256 # Test ID
+
+ @H5ZFuncT
+ def failing_filter_callback(flags, cd_nelemts, cd_values, nbytes,
buf_size, buf):
+ return 0
+
+ dummy_filter_class = H5ZClass2T(
+ version=h5z.CLASS_T_VERS,
+ id_=filter_id,
+ encoder_present=1,
+ decoder_present=1,
+ name=b"dummy filter",
+ can_apply=None,
+ set_local=None,
+ filter_=failing_filter_callback,
+ )
+
+ h5z.register_filter(addressof(dummy_filter_class))
+
+ try:
+ assert h5z.filter_avail(filter_id)
+ filter_flags = h5z.get_filter_info(filter_id)
+ assert (
+ filter_flags
+ == h5z.FILTER_CONFIG_ENCODE_ENABLED |
h5z.FILTER_CONFIG_DECODE_ENABLED
+ )
+ finally:
+ h5z.unregister_filter(filter_id)
+
+ assert not h5z.filter_avail(filter_id)
+
+
+@pytest.mark.mpi_skip
+@insubprocess
+def test_unregister_filter(request):
+ if h5py.h5z.filter_avail(h5py.h5z.FILTER_LZF):
+ res = h5py.h5z.unregister_filter(h5py.h5z.FILTER_LZF)
+ assert res
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py/version.py
new/h5py-3.9.0/h5py/version.py
--- old/h5py-3.8.0/h5py/version.py 2023-01-23 11:18:46.000000000 +0100
+++ new/h5py-3.9.0/h5py/version.py 2023-06-20 10:56:46.000000000 +0200
@@ -23,7 +23,7 @@
hdf5_built_version_tuple = _h5.HDF5_VERSION_COMPILED_AGAINST
-version_tuple = _H5PY_VERSION_CLS(3, 8, 0, None, None, None)
+version_tuple = _H5PY_VERSION_CLS(3, 9, 0, None, None, None)
version = "{0.major:d}.{0.minor:d}.{0.bugfix:d}".format(version_tuple)
if version_tuple.pre is not None:
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py.egg-info/PKG-INFO
new/h5py-3.9.0/h5py.egg-info/PKG-INFO
--- old/h5py-3.8.0/h5py.egg-info/PKG-INFO 2023-01-23 11:21:35.000000000
+0100
+++ new/h5py-3.9.0/h5py.egg-info/PKG-INFO 2023-06-20 11:06:12.000000000
+0200
@@ -1,6 +1,6 @@
Metadata-Version: 2.1
Name: h5py
-Version: 3.8.0
+Version: 3.9.0
Summary: Read and write HDF5 files from Python
Author-email: Andrew Collette <andrew.colle...@gmail.com>
Maintainer-email: Thomas Kluyver <tho...@kluyver.me.uk>, Thomas A Caswell
<tcasw...@bnl.gov>
@@ -26,7 +26,7 @@
Classifier: Topic :: Scientific/Engineering
Classifier: Topic :: Database
Classifier: Topic :: Software Development :: Libraries :: Python Modules
-Requires-Python: >=3.7
+Requires-Python: >=3.8
Description-Content-Type: text/x-rst
License-File: LICENSE
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py.egg-info/SOURCES.txt
new/h5py-3.9.0/h5py.egg-info/SOURCES.txt
--- old/h5py-3.8.0/h5py.egg-info/SOURCES.txt 2023-01-23 11:21:35.000000000
+0100
+++ new/h5py-3.9.0/h5py.egg-info/SOURCES.txt 2023-06-20 11:06:12.000000000
+0200
@@ -1,4 +1,3 @@
-.gitignore
LICENSE
MANIFEST.in
README.rst
@@ -60,6 +59,7 @@
docs/whatsnew/3.6.rst
docs/whatsnew/3.7.rst
docs/whatsnew/3.8.rst
+docs/whatsnew/3.9.rst
docs/whatsnew/index.rst
docs_api/Makefile
docs_api/automod.py
@@ -202,6 +202,7 @@
h5py/tests/test_h5p.py
h5py/tests/test_h5pl.py
h5py/tests/test_h5t.py
+h5py/tests/test_h5z.py
h5py/tests/test_objects.py
h5py/tests/test_ros3.py
h5py/tests/test_selections.py
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/h5py.egg-info/requires.txt
new/h5py-3.9.0/h5py.egg-info/requires.txt
--- old/h5py-3.8.0/h5py.egg-info/requires.txt 2023-01-23 11:21:35.000000000
+0100
+++ new/h5py-3.9.0/h5py.egg-info/requires.txt 2023-06-20 11:06:12.000000000
+0200
@@ -1 +1 @@
-numpy>=1.14.5
+numpy>=1.17.3
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/pyproject.toml
new/h5py-3.9.0/pyproject.toml
--- old/h5py-3.8.0/pyproject.toml 2023-01-20 17:24:30.000000000 +0100
+++ new/h5py-3.9.0/pyproject.toml 2023-06-07 13:39:16.000000000 +0200
@@ -1,6 +1,6 @@
[build-system]
requires = [
- "Cython ~=0.29",
+ "Cython >=0.29.31,<1",
"oldest-supported-numpy",
"pkgconfig",
"setuptools >=61",
@@ -36,7 +36,7 @@
"Topic :: Database",
"Topic :: Software Development :: Libraries :: Python Modules",
]
-requires-python = ">=3.7"
+requires-python = ">=3.8"
dynamic = ["dependencies", "version"]
[project.readme]
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/setup.py new/h5py-3.9.0/setup.py
--- old/h5py-3.8.0/setup.py 2023-01-23 11:18:46.000000000 +0100
+++ new/h5py-3.9.0/setup.py 2023-06-20 10:56:46.000000000 +0200
@@ -20,18 +20,18 @@
import setup_build, setup_configure
-VERSION = '3.8.0'
+VERSION = '3.9.0'
# these are required to use h5py
RUN_REQUIRES = [
# We only really aim to support NumPy & Python combinations for which
- # there are wheels on PyPI (e.g. NumPy >=1.17.5 for Python 3.8).
+ # there are wheels on PyPI (e.g. NumPy >=1.23.2 for Python 3.11).
# But we don't want to duplicate the information in oldest-supported-numpy
# here, and if you can build an older NumPy on a newer Python, h5py
probably
# works (assuming you build it from source too).
- # NumPy 1.14.5 is the first with wheels for Python 3.7, our minimum Python.
- "numpy >=1.14.5",
+ # NumPy 1.17.3 is the first with wheels for Python 3.8, our minimum Python.
+ "numpy >=1.17.3",
]
# Packages needed to build h5py (in addition to static list in pyproject.toml)
@@ -43,10 +43,10 @@
SETUP_REQUIRES = []
if setup_configure.mpi_enabled():
- RUN_REQUIRES.append('mpi4py >=3.0.2')
- SETUP_REQUIRES.append("mpi4py ==3.0.2; python_version<'3.8'")
- SETUP_REQUIRES.append("mpi4py ==3.0.3; python_version=='3.8.*'")
- SETUP_REQUIRES.append("mpi4py ==3.1.0; python_version=='3.9.*' or
python_version=='3.10.*'")
+ # mpi4py 3.1.1 fixed a typo in python_requires, which made older versions
+ # incompatible with newer setuptools.
+ RUN_REQUIRES.append('mpi4py >=3.1.1')
+ SETUP_REQUIRES.append("mpi4py ==3.1.1; python_version<'3.11'")
SETUP_REQUIRES.append("mpi4py ==3.1.4; python_version>='3.11'")
# Set the environment variable H5PY_SETUP_REQUIRES=0 if we need to skip
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/setup_configure.py
new/h5py-3.9.0/setup_configure.py
--- old/h5py-3.8.0/setup_configure.py 2023-01-04 18:26:33.000000000 +0100
+++ new/h5py-3.9.0/setup_configure.py 2023-06-07 13:39:16.000000000 +0200
@@ -288,6 +288,7 @@
lib = ctypes.CDLL(path, **load_kw)
except Exception:
print("error: Unable to load dependency HDF5, make sure HDF5 is
installed properly")
+ print("Library dirs checked:", libdirs)
raise
self._lib = lib
diff -urN '--exclude=CVS' '--exclude=.cvsignore' '--exclude=.svn'
'--exclude=.svnignore' old/h5py-3.8.0/tox.ini new/h5py-3.9.0/tox.ini
--- old/h5py-3.8.0/tox.ini 2023-01-20 17:24:30.000000000 +0100
+++ new/h5py-3.9.0/tox.ini 2023-06-07 13:39:16.000000000 +0200
@@ -2,7 +2,7 @@
# We want an envlist like
# envlist =
{py36,py37,pypy3}-{test}-{deps,mindeps}-{,mpi4py}-{,pre},nightly,docs,checkreadme,pre-commit
# but we want to skip mpi and pre by default, so this envlist is below
-envlist =
{py37,py38,py39,py310,py311,pypy3}-{test}-{deps,mindeps},nightly,docs,apidocs,checkreadme,pre-commit,rever
+envlist =
{py38,py39,py310,py311,pypy3}-{test}-{deps,mindeps},nightly,docs,apidocs,checkreadme,pre-commit,rever
isolated_build = True
[testenv]
@@ -11,7 +11,6 @@
test: pytest-cov
test: pytest-mpi>=0.2
- py37-deps: numpy>=1.14.5
py38-deps: numpy>=1.17.5
py39-deps: numpy>=1.19.3
py310-deps: numpy>=1.21.3
@@ -19,7 +18,7 @@
mindeps: oldest-supported-numpy
- mpi4py: mpi4py>=3.0.2
+ mpi4py: mpi4py>=3.1.1
tables-deps: tables>=3.4.4
tables-mindeps: tables==3.4.4
