Support numpy and apply fixes along the way (#26)

# v2.4.0-rc.1

- allow numpy 2.x versions
- extend support for python 3.13
- upgrade pint to 0.25
- upgrade pint-xarray up to 0.6.0
- upgrading to ontolutils 0.21.1
- fixing linting issues
- RDF-subjects set for dataset and group are modelled with property schema:about to express that a dataset or group is
  described by the RDF subject. This is more in line with semantic web standards.

Author: matthiasprobst

.github/workflows/tests.yml

@@ -6,15 +6,15 @@ on:
   push:
     branches:
       - main
-      - dev
+      - develop
 
 jobs:
   test:
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
         os: [ ubuntu-latest, macos-latest, windows-latest ]
-        python-version: [ '3.9', '3.12']
+        python-version: [ '3.9', '3.10', '3.12', '3.13' ]
         mongodb-version: ['5.0', ]
 
     steps:

CHANGELOG.md

@@ -2,11 +2,21 @@
 
 Log of changes in the versions
 
+## v2.4.0-rc.1
+
+- allow numpy 2.x versions
+- extend support for python 3.13
+- upgrade pint to 0.25
+- upgrade pint-xarray up to 0.6.0
+- upgrading to ontolutils 0.21.1
+- fixing linting issues
+- RDF-subjects set for dataset and group are modelled with property schema:about to express that a dataset or group is
+  described by the RDF subject. This is more in line with semantic web standards.
+
 ## v2.3.1
 
 - fixing error in parsing obj name. "/" is safed and will not be converted anymore
 
-
 ## v2.3.0
 
 - hotfix avoiding blank nodes for hdf filter

CITATION.cff

@@ -11,7 +11,7 @@ authors:
     given-names: "Lucas"
     orcid: "https://orcid.org/0000-0002-4116-0065"
 title: "h5rdmtoolbox - HDF5 Research Data Management Toolbox"
-version: 2.3.1
-doi: 10.5281/zenodo.17063344
-date-released: 2025-09-01
+version: 2.4.0-rc.1
+doi: 10.5281/zenodo.17088483
+date-released: 2025-10-12
 url: "https://github.com/matthiasprobst/h5rdmtoolbox"

README.md

@@ -3,7 +3,7 @@
 ![Tests](https://github.com/matthiasprobst/h5RDMtoolbox/actions/workflows/tests.yml/badge.svg)
 [![codecov](https://codecov.io/gh/matthiasprobst/h5RDMtoolbox/graph/badge.svg?token=IVG4AQEW47)](https://codecov.io/gh/matthiasprobst/h5RDMtoolbox)
 [![Documentation Status](https://readthedocs.org/projects/h5rdmtoolbox/badge/?version=latest)](https://h5rdmtoolbox.readthedocs.io/en/latest/?badge=latest)
-![pyvers](https://img.shields.io/badge/python-3.9%20%7C%203.10%20%7C%203.11%20%7C%203.12-blue)
+![pyvers](https://img.shields.io/badge/python-3.9%20%7C%203.10%20%7C%203.11%20%7C%203.12%20%7C%203.13-blue)
 
 *Note, that the project is still under development!*
 
@@ -99,6 +99,10 @@ package via pip:
 
     pip install h5RDMtoolbox
 
+or if you prefer `uv`:
+
+    uv install h5RDMtoolbox
+
 ### Install from source:
 
 Developers may clone the repository and install the package from source. Clone the repository first:
@@ -131,7 +135,7 @@ specific to the features of the package:
 
 **Specific to the package are ...**
 
-- `h5py=3.7.0`: HDF5 file interface
+- `h5py~=3.8`: HDF5 file interface
 - `xarray>=2022.3.0`: Working with scientific arrays in combination with attributes. Allows carrying metadata from HDF5
   to user
 - `pint>=0.19.2`: Allows working with units

codemeta.json

@@ -4,7 +4,7 @@
     "license": "https://spdx.org/licenses/MIT",
     "codeRepository": "git+https://github.com/matthiasprobst/h5RDMtoolbox.git",
     "name": "h5RDMtoolbox",
-    "version": "2.3.1",
+    "version": "2.4.0-rc.1",
     "description": "Supporting a FAIR Research Data lifecycle using Python and HDF5.",
     "applicationCategory": "Engineering",
     "programmingLanguage": [

docs/colab/quickstart.ipynb

@@ -7,7 +7,7 @@
     "scrolled": true
    },
    "outputs": [],
-   "source": "# !pip install h5rdmtoolbox==2.3.1"
+   "source": "# !pip install h5rdmtoolbox==2.4.0rc1"
   },
   {
    "cell_type": "code",

docs/gettingstarted/quickoverview.ipynb

@@ -62,7 +62,7 @@
 __author_orcid__ = 'https://orcid.org/0000-0001-8729-0482'
 
 
-def get_package_meta():
+def get_package_meta() -> Dict:
     """Reads codemeta.json and returns it as dict"""
     with open(__this_dir__ / '../codemeta.json', 'r') as f:
         codemeta = json.loads(f.read())
@@ -172,7 +172,6 @@ def dump_jsonld_depr(hdf_filename: Union[str, pathlib.Path],
                      skipND: int = 1,
                      structural: bool = True,
                      contextual: bool = True,
-                     resolve_keys: bool = True,
                      context: Optional[Dict] = None,
                      blank_node_iri_base: Optional[str] = None
                      ) -> str:
@@ -189,19 +188,12 @@ def dump_jsonld_depr(hdf_filename: Union[str, pathlib.Path],
         considered if structural=True.
     structural : bool=True
         Include structural information in the JSON-LD output.
-    semantic : bool=True
-        Include semantic information in the JSON-LD output.
-    resolve_keys : bool=True
-        Resolve keys in the JSON-LD output. This is used when semantic=True.
-        If resolve_keys is False and an attribute name in the HDF5 file, which has
-        a predicate is different in its name from the predicate, the attribute name is used.
-        Example: an attribute "name" is associated with "foaf:lastName", then "name" is used
-        and "name": "http://xmlns.com/foaf/0.1/lastName" is added to the context.
+    contextual : bool=True
+        Include contextual information in the JSON-LD output.
     context: Optional[Dict]
         context in form of {prefix: IRI}, e.g. "ssno": "https://matthiasprobst.github.io/ssno#"
-        If resolve_keys is True, this is added to the built-in look-up table to be used in the
-        context part of JSON-LD
-
+    blank_node_iri_base: Optional[str]
+        IRI base used for blank nodes
     """
     if blank_node_iri_base is not None:
         if not isinstance(blank_node_iri_base, (str, HttpUrl)):
@@ -212,20 +204,18 @@ def dump_jsonld_depr(hdf_filename: Union[str, pathlib.Path],
 
     from h5rdmtoolbox.ld.hdf.file import get_serialized_ld
     if structural and not contextual:
-        return get_serialized_ld(hdf_filename, blank_node_iri_base, format="json-ld", context=context,
-                                 contextual=contextual, skipND=skipND)
-    return get_serialized_ld(hdf_filename, blank_node_iri_base, format="json-ld", context=context,
-                             contextual=contextual, skipND=skipND)
-    # with File(hdf_filename) as h5:
-    #     return jsonld.dumps(
-    #         h5,
-    #         structural=structural,
-    #         resolve_keys=resolve_keys,
-    #         context=context,
-    #         blank_node_iri_base=blank_node_iri_base,
-    #         skipND=skipND,
-    #         **kwargs
-    #     )
+        return get_serialized_ld(
+            hdf_filename,
+            blank_node_iri_base,
+            format="json-ld",
+            context=context,
+            skipND=skipND)
+    return get_serialized_ld(
+        hdf_filename,
+        blank_node_iri_base,
+        format="json-ld",
+        context=context,
+        skipND=skipND)
 
 
 def serialize(hdf_filename,
@@ -257,7 +247,7 @@ def build_pyvis_graph(hdf_filename, output_filename="kg-graph.html", notebook=Fa
     kg = kglab.KnowledgeGraph().load_rdf_text(
         serialize(hdf_filename, fmt="ttl", structural=structural, contextual=contextual)
     )
-    VIS_STYLE = style or {
+    vis_style = style or {
         "hdf": {
             "color": "orange",
             "size": 40,
@@ -268,7 +258,7 @@ def build_pyvis_graph(hdf_filename, output_filename="kg-graph.html", notebook=Fa
         },
     }
     subgraph = kglab.SubgraphTensor(kg)
-    pyvis_graph = subgraph.build_pyvis_graph(notebook=notebook, style=VIS_STYLE)
+    pyvis_graph = subgraph.build_pyvis_graph(notebook=notebook, style=vis_style)
     pyvis_graph.force_atlas_2based()
     pyvis_graph.show(output_filename)
     return pyvis_graph
@@ -372,4 +362,5 @@ def clean_temp_data(full: bool = False):
            'File', 'Group', 'Dataset', 'Attribute',
            'dump', 'dumps', 'cv_h5py', 'lower', 'Lower',
            'set_config', 'get_config', 'get_ureg',
-           'Convention', 'jsonld', 'lazy', 'DownloadFileManager')
+           'Convention', 'jsonld', 'lazy', 'DownloadFileManager',
+           'get_package_meta', 'clean_temp_data')

h5rdmtoolbox/__init__.py

@@ -73,7 +73,7 @@ def __init__(self, **kwargs):
                 raise KeyError(f'Not a configuration key: "{k}"')
             self.old[k] = CONFIG[k]
             if k == 'ureg_format':
-                get_ureg().default_format = str(v)
+                get_ureg().formatter.default_format = str(v)
         self._update(kwargs)
 
     def __enter__(self):

h5rdmtoolbox/_cfg.py

@@ -13,6 +13,8 @@
 from pydoc import locate
 from typing import Union, List, Dict, Tuple, Any
 
+from forge import kwargs
+
 from h5rdmtoolbox import errors
 from h5rdmtoolbox.repository import RepositoryInterface
 from h5rdmtoolbox.wrapper import ds_decoder
@@ -696,24 +698,19 @@ def from_json(filename: Union[str, pathlib.Path], overwrite: bool = False) -> Co
 
 
 def from_repo(repo_interface: RepositoryInterface,
-              name: str,
-              take_existing: bool = True,
-              force_download: bool = False):
-    """Download a YAML file from a repository"""
-    # check if file exists:
-    # path_compatible_doi = repo_interface.get_doi().replace('/', '_')
-    # estimated_filename = UserDir['cache'] / f'{path_compatible_doi}' / name
-    # estimated_filename.parent.mkdir(parents=True, exist_ok=True)
-    # if estimated_filename.exists():
-    #     if not take_existing:
-    #         raise FileExistsError(f'File {name} exists in cache but take_existing is set to False.')
-    #     if take_existing and not force_download:
-    #         return from_file(estimated_filename)
+              name: str):
+    """Download a YAML file from a repository
 
+    Parameters
+    ----------
+    repo_interface: RepositoryInterface
+        The repository interface to use for downloading the file
+    name: str
+        Name of the file to download
+    """
+    logger.debug(f"Downloading file {name} from repository {repo_interface}")
     filename = repo_interface.download_file(name)
-    # if estimated_filename.exists():
-    #     estimated_filename.unlink()
-    # filename.rename(estimated_filename)
+    logger.debug(f"File downloaded to {filename}. Now loading convention from file.")
     return from_file(filename)
 
 
@@ -723,7 +720,7 @@ def from_zenodo(doi_or_recid: str,
                 force_download: bool = False) -> Convention:
     """Download a YAML file from a zenodo repository
 
-    Depreciated. Use `from_repo` in future.
+    Depreciated. Use `from_repo` in the future.
 
     Parameters
     ----------