cognite-neat 0.96.5__tar.gz → 0.97.0__tar.gz

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: cognite-neat
-Version: 0.96.5
+Version: 0.97.0
 Summary: Knowledge graph transformation
 Home-page: https://cognite-neat.readthedocs-hosted.com/
 License: Apache-2.0
@@ -16,18 +16,19 @@ Classifier: Programming Language :: Python :: 3.13
 Provides-Extra: all
 Provides-Extra: docs
 Provides-Extra: google
-Provides-Extra: graphql
 Provides-Extra: oxi
 Provides-Extra: service
 Requires-Dist: PyYAML
 Requires-Dist: backports.strenum (>=1.2,<2.0) ; python_version < "3.11"
 Requires-Dist: cognite-sdk (>=7.54.6,<8.0.0)
+Requires-Dist: elementpath (>=4.0.0,<5.0.0)
 Requires-Dist: exceptiongroup (>=1.1.3,<2.0.0) ; python_version < "3.11"
 Requires-Dist: fastapi (>=0,<1) ; extra == "service" or extra == "all"
 Requires-Dist: google-api-python-client ; extra == "google"
 Requires-Dist: google-auth-oauthlib ; extra == "google"
 Requires-Dist: gspread ; extra == "google"
-Requires-Dist: jinja2 (>=3.1.2,<4.0.0) ; extra == "graphql" or extra == "all"
+Requires-Dist: jsonpath-python (>=1.0.6,<2.0.0)
+Requires-Dist: lxml (>=5.3.0,<6.0.0)
 Requires-Dist: mkdocs ; extra == "docs"
 Requires-Dist: mkdocs-autorefs (>=0.5.0,<0.6.0) ; extra == "docs"
 Requires-Dist: mkdocs-git-authors-plugin ; extra == "docs"
@@ -40,6 +41,7 @@ Requires-Dist: mkdocstrings[python] ; extra == "docs"
 Requires-Dist: networkx (>=3.4.2,<4.0.0)
 Requires-Dist: openpyxl
 Requires-Dist: oxrdflib[oxigraph] (>=0.3.3,<0.4.0) ; extra == "oxi" or extra == "all"
+Requires-Dist: packaging (>=22.0,<25.0)
 Requires-Dist: pandas
 Requires-Dist: prometheus-client (>=0,<1) ; extra == "service" or extra == "all"
 Requires-Dist: pydantic (>=2,<3)

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_constants.py

@@ -36,6 +36,7 @@ COGNITE_MODELS = (
     DataModelId("cdf_cdm", "CogniteCore", "v1"),
     DataModelId("cdf_idm", "CogniteProcessIndustries", "v1"),
 )
+COGNITE_SPACES = frozenset(model.space for model in COGNITE_MODELS)
 DMS_LISTABLE_PROPERTY_LIMIT = 1000
 
 EXAMPLE_RULES = PACKAGE_DIRECTORY / "_rules" / "examples"
@@ -43,10 +44,12 @@ EXAMPLE_GRAPHS = PACKAGE_DIRECTORY / "_graph" / "examples"
 EXAMPLE_WORKFLOWS = PACKAGE_DIRECTORY / "_workflows" / "examples"
 
 DEFAULT_SPACE_URI = "http://purl.org/cognite/{space}#"
-DEFAULT_NAMESPACE = Namespace("http://purl.org/cognite/neat#")
+DEFAULT_NAMESPACE = Namespace("http://purl.org/cognite/neat/")
+CDF_NAMESPACE = Namespace("https://cognitedata.com/")
 DEFAULT_BASE_URI = URIRef(DEFAULT_NAMESPACE)
 CLASSIC_CDF_NAMESPACE = Namespace("http://purl.org/cognite/cdf-classic#")
 UNKNOWN_TYPE = DEFAULT_NAMESPACE.UnknownType
+XML_SCHEMA_NAMESPACE = Namespace("http://www.w3.org/2001/XMLSchema#")
 
 
 def get_default_prefixes() -> dict[str, Namespace]:

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_graph/extractors/__init__.py

@@ -1,3 +1,5 @@
+from cognite.neat._session.engine._interface import Extractor as EngineExtractor
+
 from ._base import BaseExtractor
 from ._classic_cdf._assets import AssetsExtractor
 from ._classic_cdf._classic import ClassicGraphExtractor
@@ -48,6 +50,7 @@ TripleExtractors = (
     | DMSExtractor
     | ClassicGraphExtractor
     | DataSetExtractor
+    | EngineExtractor
 )
 
 

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_graph/extractors/_base.py

@@ -1,7 +1,7 @@
 from abc import abstractmethod
 from collections.abc import Iterable
 
-from cognite.neat._graph.models import Triple
+from cognite.neat._shared import Triple
 from cognite.neat._utils.auxiliary import class_html_doc
 
 

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_graph/extractors/_classic_cdf/_assets.py

@@ -7,7 +7,7 @@ from cognite.client import CogniteClient
 from cognite.client.data_classes import Asset, AssetFilter, AssetList
 from rdflib import RDF, Literal, Namespace
 
-from cognite.neat._graph.models import Triple
+from cognite.neat._shared import Triple
 
 from ._base import DEFAULT_SKIP_METADATA_VALUES, ClassicCDFBaseExtractor, InstanceIdPrefix
 from ._labels import LabelsExtractor

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_graph/extractors/_classic_cdf/_base.py

@@ -10,7 +10,7 @@ from rdflib import XSD, Literal, Namespace, URIRef
 
 from cognite.neat._constants import DEFAULT_NAMESPACE
 from cognite.neat._graph.extractors._base import BaseExtractor
-from cognite.neat._graph.models import Triple
+from cognite.neat._shared import Triple
 from cognite.neat._utils.auxiliary import string_to_ideal_type
 
 T_CogniteResource = TypeVar("T_CogniteResource", bound=CogniteResource)

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_graph/extractors/_classic_cdf/_classic.py

@@ -7,7 +7,7 @@ from rdflib import Namespace
 
 from cognite.neat._constants import DEFAULT_NAMESPACE
 from cognite.neat._graph.extractors._base import BaseExtractor
-from cognite.neat._graph.models import Triple
+from cognite.neat._shared import Triple
 from cognite.neat._utils.collection_ import chunker
 from cognite.neat._utils.rdf_ import remove_namespace_from_uri
 

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_graph/extractors/_classic_cdf/_data_sets.py

@@ -7,7 +7,7 @@ from cognite.client.data_classes import DataSet, DataSetList
 from cognite.client.utils.useful_types import SequenceNotStr
 from rdflib import RDF, Literal, Namespace
 
-from cognite.neat._graph.models import Triple
+from cognite.neat._shared import Triple
 
 from ._base import DEFAULT_SKIP_METADATA_VALUES, ClassicCDFBaseExtractor, InstanceIdPrefix
 

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_graph/extractors/_classic_cdf/_events.py

@@ -6,7 +6,7 @@ from cognite.client import CogniteClient
 from cognite.client.data_classes import Event, EventFilter, EventList
 from rdflib import RDF, Literal, Namespace
 
-from cognite.neat._graph.models import Triple
+from cognite.neat._shared import Triple
 
 from ._base import DEFAULT_SKIP_METADATA_VALUES, ClassicCDFBaseExtractor, InstanceIdPrefix
 

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_graph/extractors/_classic_cdf/_files.py

@@ -6,7 +6,7 @@ from cognite.client import CogniteClient
 from cognite.client.data_classes import FileMetadata, FileMetadataFilter, FileMetadataList
 from rdflib import RDF, Literal, Namespace
 
-from cognite.neat._graph.models import Triple
+from cognite.neat._shared import Triple
 
 from ._base import DEFAULT_SKIP_METADATA_VALUES, ClassicCDFBaseExtractor, InstanceIdPrefix
 from ._labels import LabelsExtractor

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_graph/extractors/_classic_cdf/_labels.py

@@ -7,7 +7,7 @@ from cognite.client import CogniteClient
 from cognite.client.data_classes import Label, LabelDefinition, LabelDefinitionList
 from rdflib import RDF, Literal, Namespace
 
-from cognite.neat._graph.models import Triple
+from cognite.neat._shared import Triple
 
 from ._base import DEFAULT_SKIP_METADATA_VALUES, ClassicCDFBaseExtractor, InstanceIdPrefix
 

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_graph/extractors/_classic_cdf/_relationships.py

@@ -7,7 +7,7 @@ from cognite.client import CogniteClient
 from cognite.client.data_classes import Relationship, RelationshipList
 from rdflib import RDF, Literal, Namespace
 
-from cognite.neat._graph.models import Triple
+from cognite.neat._shared import Triple
 from cognite.neat._utils.auxiliary import create_sha256_hash
 
 from ._base import DEFAULT_SKIP_METADATA_VALUES, ClassicCDFBaseExtractor, InstanceIdPrefix

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_graph/extractors/_classic_cdf/_sequences.py

@@ -6,7 +6,7 @@ from cognite.client import CogniteClient
 from cognite.client.data_classes import Sequence, SequenceFilter, SequenceList
 from rdflib import RDF, Literal, Namespace
 
-from cognite.neat._graph.models import Triple
+from cognite.neat._shared import Triple
 
 from ._base import DEFAULT_SKIP_METADATA_VALUES, ClassicCDFBaseExtractor, InstanceIdPrefix
 

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_graph/extractors/_classic_cdf/_timeseries.py

@@ -7,7 +7,7 @@ from cognite.client.data_classes import TimeSeries, TimeSeriesFilter, TimeSeries
 from pydantic import AnyHttpUrl, ValidationError
 from rdflib import RDF, Literal, Namespace, URIRef
 
-from cognite.neat._graph.models import Triple
+from cognite.neat._shared import Triple
 
 from ._base import DEFAULT_SKIP_METADATA_VALUES, ClassicCDFBaseExtractor, InstanceIdPrefix
 

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_graph/extractors/_dexpi.py

@@ -8,7 +8,7 @@ from rdflib import RDF, RDFS, XSD, Literal, Namespace, URIRef
 
 from cognite.neat._constants import DEFAULT_NAMESPACE
 from cognite.neat._graph.extractors._base import BaseExtractor
-from cognite.neat._graph.models import Triple
+from cognite.neat._shared import Triple
 from cognite.neat._utils.rdf_ import as_neat_compliant_uri
 from cognite.neat._utils.xml_ import get_children, iterate_tree
 

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_graph/extractors/_dms.py

@@ -8,8 +8,8 @@ from cognite.client.data_classes.data_modeling.instances import Instance, Proper
 from rdflib import RDF, XSD, Literal, Namespace, URIRef
 
 from cognite.neat._constants import DEFAULT_SPACE_URI
-from cognite.neat._graph.models import Triple
 from cognite.neat._issues.errors import ResourceRetrievalError
+from cognite.neat._shared import Triple
 
 from ._base import BaseExtractor
 

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_graph/extractors/_iodd.py

@@ -11,8 +11,8 @@ from rdflib import RDF, XSD, Literal, Namespace, URIRef
 
 from cognite.neat._constants import DEFAULT_NAMESPACE
 from cognite.neat._graph.extractors._base import BaseExtractor
-from cognite.neat._graph.models import Triple
 from cognite.neat._issues.errors import FileReadError, NeatValueError
+from cognite.neat._shared import Triple
 from cognite.neat._utils.rdf_ import remove_namespace_from_uri
 from cognite.neat._utils.text import to_camel
 from cognite.neat._utils.xml_ import get_children

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_graph/extractors/_mock_graph_generator.py

@@ -11,13 +11,13 @@ import numpy
 import pandas as pd
 from rdflib import RDF, Literal, Namespace, URIRef
 
-from cognite.neat._graph.models import Triple
 from cognite.neat._rules._constants import EntityTypes
 from cognite.neat._rules.analysis import InformationAnalysis
 from cognite.neat._rules.models import DMSRules, InformationRules
 from cognite.neat._rules.models.data_types import DataType
 from cognite.neat._rules.models.entities import ClassEntity
 from cognite.neat._rules.models.information import InformationProperty
+from cognite.neat._shared import Triple
 from cognite.neat._utils.rdf_ import remove_namespace_from_uri
 
 from ._base import BaseExtractor

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_graph/extractors/_rdf_file.py

@@ -8,9 +8,9 @@ from rdflib.util import guess_format
 from cognite.neat._constants import DEFAULT_BASE_URI
 from cognite.neat._graph._shared import rdflib_to_mime_types
 from cognite.neat._graph.extractors._base import BaseExtractor
-from cognite.neat._graph.models import Triple
 from cognite.neat._issues._base import IssueList
 from cognite.neat._issues.errors import FileNotFoundNeatError, FileTypeUnexpectedError
+from cognite.neat._shared import Triple
 
 
 class RdfFileExtractor(BaseExtractor):

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_graph/loaders/_rdf2dms.py

@@ -16,7 +16,6 @@ from pydantic import BaseModel, ValidationInfo, create_model, field_validator
 from rdflib import RDF
 
 from cognite.neat._graph._tracking import LogTracker, Tracker
-from cognite.neat._graph.models import InstanceType
 from cognite.neat._issues import IssueList, NeatIssue, NeatIssueList
 from cognite.neat._issues.errors import (
     ResourceConvertionError,
@@ -27,6 +26,7 @@ from cognite.neat._issues.errors import (
 from cognite.neat._issues.warnings import PropertyTypeNotSupportedWarning
 from cognite.neat._rules.models import DMSRules
 from cognite.neat._rules.models.data_types import _DATA_TYPE_BY_DMS_TYPE, Json
+from cognite.neat._shared import InstanceType
 from cognite.neat._store import NeatGraphStore
 from cognite.neat._utils.auxiliary import create_sha256_hash
 from cognite.neat._utils.upload import UploadResult

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_graph/queries/_base.py

@@ -8,10 +8,10 @@ from rdflib import Literal as RdfLiteral
 from rdflib.query import ResultRow
 
 from cognite.neat._constants import UNKNOWN_TYPE
-from cognite.neat._graph.models import InstanceType
 from cognite.neat._rules._constants import EntityTypes
 from cognite.neat._rules.models.entities import ClassEntity
 from cognite.neat._rules.models.information import InformationRules
+from cognite.neat._shared import InstanceType
 from cognite.neat._utils.rdf_ import remove_namespace_from_uri
 
 from ._construct import build_construct_query

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_graph/transformers/__init__.py

@@ -7,7 +7,7 @@ from ._classic_cdf import (
     AssetTimeSeriesConnector,
     RelationshipToSchemaTransformer,
 )
-from ._rdfpath import AddSelfReferenceProperty
+from ._rdfpath import AddSelfReferenceProperty, MakeConnectionOnExactMatch
 from ._value_type import SplitMultiValueProperty
 
 __all__ = [
@@ -20,6 +20,7 @@ __all__ = [
     "AddSelfReferenceProperty",
     "SplitMultiValueProperty",
     "RelationshipToSchemaTransformer",
+    "MakeConnectionOnExactMatch",
 ]
 
 Transformers = (
@@ -32,4 +33,5 @@ Transformers = (
     | AddSelfReferenceProperty
     | SplitMultiValueProperty
     | RelationshipToSchemaTransformer
+    | MakeConnectionOnExactMatch
 )

cognite_neat-0.97.0/cognite/neat/_graph/transformers/_rdfpath.py

@@ -0,0 +1,109 @@
+from typing import cast
+from urllib.parse import quote
+
+from rdflib import Graph, URIRef
+
+from cognite.neat._constants import DEFAULT_NAMESPACE
+from cognite.neat._rules.analysis import InformationAnalysis
+from cognite.neat._rules.models._rdfpath import RDFPath, SingleProperty
+from cognite.neat._rules.models.information import InformationRules
+from cognite.neat._shared import Triple
+from cognite.neat._utils.rdf_ import add_triples_in_batch, remove_namespace_from_uri
+
+from ._base import BaseTransformer
+
+
+class ReduceHopTraversal(BaseTransformer):
+    """ReduceHopTraversal is a transformer that reduces the number of hops to direct connection."""
+
+    ...
+
+
+class AddSelfReferenceProperty(BaseTransformer):
+    description: str = "Adds property that contains id of reference to all references of given class in Rules"
+    _use_only_once: bool = True
+    _need_changes = frozenset({})
+    _ref_template: str = """SELECT ?s WHERE {{?s a <{type_}>}}"""
+
+    def __init__(
+        self,
+        rules: InformationRules,
+    ):
+        self.rules = rules
+        self.properties = InformationAnalysis(rules).all_reference_transformations()
+
+    def transform(self, graph: Graph) -> None:
+        for property_ in self.properties:
+            prefix = property_.transformation.traversal.class_.prefix
+            suffix = property_.transformation.traversal.class_.suffix
+
+            namespace = self.rules.prefixes[prefix] if prefix in self.rules.prefixes else self.rules.metadata.namespace
+
+            for (reference,) in graph.query(self._ref_template.format(type_=namespace[suffix])):  # type: ignore [misc]
+                graph.add(
+                    (
+                        reference,
+                        self.rules.metadata.namespace[property_.property_],
+                        reference,
+                    )
+                )
+
+            traversal = SingleProperty.from_string(
+                class_=property_.class_.id,
+                property_=f"{self.rules.metadata.prefix}:{property_.property_}",
+            )
+
+            property_.transformation = RDFPath(traversal=traversal)
+
+
+class MakeConnectionOnExactMatch(BaseTransformer):
+    description: str = "Adds property that contains id of reference to all references of given class in Rules"
+    _use_only_once: bool = True
+    _need_changes = frozenset({})
+    _ref_template: str = """SELECT DISTINCT ?subject ?object
+                            WHERE {{
+                                ?subject a <{subject_type}> .
+                                ?subject <{subject_predicate}> ?value .
+                                ?object <{object_predicate}> ?value .
+                                ?object a <{object_type}> .
+                            }}"""
+
+    def __init__(
+        self,
+        subject_type: URIRef,
+        subject_predicate: URIRef,
+        object_type: URIRef,
+        object_predicate: URIRef,
+        connection: URIRef | str | None = None,
+        limit: int | None = None,
+    ):
+        self.subject_type = subject_type
+        self.subject_predicate = subject_predicate
+        self.object_type = object_type
+        self.object_predicate = object_predicate
+
+        self.connection = (
+            DEFAULT_NAMESPACE[quote(connection.strip())]
+            if isinstance(connection, str)
+            else connection or DEFAULT_NAMESPACE[remove_namespace_from_uri(self.object_type).lower()]
+        )
+
+        self.limit = limit
+
+    def transform(self, graph: Graph) -> None:
+        query = self._ref_template.format(
+            subject_type=self.subject_type,
+            subject_predicate=self.subject_predicate,
+            object_type=self.object_type,
+            object_predicate=self.object_predicate,
+        )
+
+        if self.limit and isinstance(self.limit, int) and self.limit > 0:
+            query += f" LIMIT {self.limit}"
+
+        triples: list[Triple] = []
+        for subject, object in graph.query(query):  # type: ignore [misc]
+            triples.append(cast(Triple, (subject, self.connection, object)))
+
+        print(f"Found {len(triples)} connections. Adding them to the graph...")
+        add_triples_in_batch(graph, triples)

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_issues/errors/__init__.py

@@ -16,6 +16,7 @@ from ._properties import (
     PropertyMappingDuplicatedError,
     PropertyNotFoundError,
     PropertyTypeNotSupportedError,
+    ReversedConnectionNotFeasibleError,
 )
 from ._resources import (
     ResourceChangedError,
@@ -68,6 +69,7 @@ __all__ = [
     "PropertyMappingDuplicatedError",
     "RowError",
     "NeatTypeError",
+    "ReversedConnectionNotFeasibleError",
 ]
 
 _NEAT_ERRORS_BY_NAME = {error.__name__: error for error in _get_subclasses(NeatError, include_base=True)}

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_issues/errors/_properties.py

@@ -32,6 +32,18 @@ class PropertyTypeNotSupportedError(PropertyError[T_Identifier]):
     property_type: str
 
 
+@dataclass(unsafe_hash=True)
+class ReversedConnectionNotFeasibleError(PropertyError[T_Identifier]):
+    """The {resource_type} {property_name} with identifier {identifier} of the view {target_view_id} cannot be made
+    since view {source_view_id} does not have direct connection {direct_connection} defined,
+    or {direct_connection} value type is not {target_view_id}
+    """
+
+    target_view_id: str
+    source_view_id: str
+    direct_connection: str
+
+
 # This is a generic error that should be used sparingly
 @dataclass(unsafe_hash=True)
 class PropertyDefinitionError(PropertyError[T_Identifier]):

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_issues/warnings/__init__.py

@@ -20,6 +20,7 @@ from ._models import (
     PrincipleMatchingSpaceAndVersionWarning,
     PrincipleOneModelOneSpaceWarning,
     PrincipleSolutionBuildsOnEnterpriseWarning,
+    UndefinedViewWarning,
     UserModelingWarning,
 )
 from ._properties import (
@@ -62,6 +63,7 @@ __all__ = [
     "PrincipleSolutionBuildsOnEnterpriseWarning",
     "NotSupportedViewContainerLimitWarning",
     "NotSupportedHasDataFilterLimitWarning",
+    "UndefinedViewWarning",
     "user_modeling",
 ]
 

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_issues/warnings/_models.py

@@ -92,6 +92,17 @@ class NotSupportedHasDataFilterLimitWarning(CDFNotSupportedWarning):
     limit: int = DMS_VIEW_CONTAINER_SIZE_LIMIT
 
 
+@dataclass(unsafe_hash=True)
+class UndefinedViewWarning(UserModelingWarning):
+    """Undefined view {value_type} has been referred as value type for property <{view_property}> of view {view_id}."""
+
+    fix = "Define views which are used as value types."
+
+    view_id: str
+    value_type: str
+    view_property: str
+
+
 @dataclass(unsafe_hash=True)
 class EnterpriseModelNotBuildOnTopOfCDMWarning(UserModelingWarning):
     """Enterprise data model being build on top {reference_model_id}. This is not recommended."""

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_rules/importers/__init__.py

@@ -19,6 +19,17 @@ __all__ = [
     "InferenceImporter",
 ]
 
+RulesImporters = (
+    OWLImporter
+    | IMFImporter
+    | DMSImporter
+    | ExcelImporter
+    | GoogleSheetImporter
+    | DTDLImporter
+    | YAMLImporter
+    | InferenceImporter
+)
+
 
 def _repr_html_() -> str:
     import pandas as pd

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_rules/importers/_base.py

@@ -8,8 +8,10 @@ from typing import Any, Generic, Literal
 from pydantic import ValidationError
 from rdflib import Namespace
 
+from cognite.neat._constants import DEFAULT_NAMESPACE
 from cognite.neat._issues import IssueList, NeatError, NeatWarning
 from cognite.neat._rules._shared import ReadRules, T_InputRules
+from cognite.neat._store._provenance import Agent as ProvenanceAgent
 from cognite.neat._utils.auxiliary import class_html_doc
 
 
@@ -46,6 +48,11 @@ class BaseImporter(ABC, Generic[T_InputRules]):
     def _repr_html_(cls) -> str:
         return class_html_doc(cls)
 
+    @property
+    def agent(self) -> ProvenanceAgent:
+        """Provenance agent for the importer."""
+        return ProvenanceAgent(id_=DEFAULT_NAMESPACE[f"agent/{type(self).__name__}"])
+
 
 class _FutureResult:
     def __init__(self) -> None:

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_rules/importers/_dms2rules.py

@@ -1,6 +1,6 @@
 from collections import Counter
 from collections.abc import Collection, Sequence
-from datetime import datetime
+from datetime import datetime, timezone
 from pathlib import Path
 from typing import Literal, cast
 
@@ -104,6 +104,7 @@ class DMSImporter(BaseImporter[DMSInputRules]):
         Returns:
             DMSImporter: DMSImporter instance
         """
+
         data_model_ids = [data_model_id, reference_model_id] if reference_model_id else [data_model_id]
         data_models = client.data_modeling.data_models.retrieve(data_model_ids, inline_views=True)
 
@@ -128,7 +129,9 @@ class DMSImporter(BaseImporter[DMSInputRules]):
                     DMSSchema(),
                     [
                         ResourceRetrievalError(
-                            dm.DataModelId.load(reference_model_id), "data model", "Data Model is missing in CDF"
+                            dm.DataModelId.load(reference_model_id),
+                            "data model",
+                            "Data Model is missing in CDF",
                         )
                     ],
                 )
@@ -201,7 +204,10 @@ class DMSImporter(BaseImporter[DMSInputRules]):
     @classmethod
     def from_zip_file(cls, zip_file: str | Path) -> "DMSImporter":
         if Path(zip_file).suffix != ".zip":
-            return cls(DMSSchema(), [FileTypeUnexpectedError(Path(zip_file), frozenset([".zip"]))])
+            return cls(
+                DMSSchema(),
+                [FileTypeUnexpectedError(Path(zip_file), frozenset([".zip"]))],
+            )
         issue_list = IssueList()
         with _handle_issues(issue_list) as _:
             schema = DMSSchema.from_zip(zip_file)
@@ -210,10 +216,12 @@ class DMSImporter(BaseImporter[DMSInputRules]):
     def to_rules(self) -> ReadRules[DMSInputRules]:
         if self.issue_list.has_errors:
             # In case there were errors during the import, the to_rules method will return None
+            self._end = datetime.now(timezone.utc)
             return ReadRules(None, self.issue_list, {})
 
         if not self.root_schema.data_model:
             self.issue_list.append(ResourceMissingIdentifierError("data model", type(self.root_schema).__name__))
+            self._end = datetime.now(timezone.utc)
             return ReadRules(None, self.issue_list, {})
 
         model = self.root_schema.data_model
@@ -240,6 +248,7 @@ class DMSImporter(BaseImporter[DMSInputRules]):
             has_reference=reference is not None,
         )
         user_rules.reference = reference
+        self._end = datetime.now(timezone.utc)
 
         return ReadRules(user_rules, self.issue_list, {})
 

{cognite_neat-0.96.5 → cognite_neat-0.97.0}/cognite/neat/_rules/importers/_rdf/_inference2rules.py

@@ -1,13 +1,15 @@
 from collections import Counter, defaultdict
+from collections.abc import Mapping
 from datetime import datetime
 from pathlib import Path
-from typing import cast
+from typing import ClassVar, cast
 
-from rdflib import RDF
+from rdflib import RDF, URIRef
 from rdflib import Literal as RdfLiteral
 
 from cognite.neat._constants import DEFAULT_NAMESPACE
 from cognite.neat._issues.warnings import PropertyValueTypeUndefinedWarning
+from cognite.neat._rules.models import data_types
 from cognite.neat._rules.models._base_rules import MatchType
 from cognite.neat._rules.models.data_types import AnyURI
 from cognite.neat._rules.models.entities._single_value import UnknownEntity
@@ -41,13 +43,24 @@ class InferenceImporter(BaseRDFImporter):
     Use the factory methods to create a triple store from sources such as
     RDF files, JSON files, YAML files, XML files, or directly from a graph store.
 
+    ClassVars:
+        overwrite_data_types: Mapping of data types to be overwritten. The InferenceImporter will overwrite
+            32-bit integer and 32-bit float data types to 64-bit integer and 64-bit float data types
+
     Args:
         issue_list: Issue list to store issues
         graph: Knowledge graph
         max_number_of_instance: Maximum number of instances to be used in inference
         prefix: Prefix to be used for the inferred model
+
+
     """
 
+    overwrite_data_types: ClassVar[Mapping[URIRef, URIRef]] = {
+        data_types.Integer.as_xml_uri_ref(): data_types.Long.as_xml_uri_ref(),
+        data_types.Float.as_xml_uri_ref(): data_types.Double.as_xml_uri_ref(),
+    }
+
     @classmethod
     def from_graph_store(
         cls,
@@ -140,6 +153,8 @@ class InferenceImporter(BaseRDFImporter):
                         continue
 
                     property_id = remove_namespace_from_uri(property_uri)
+                    if isinstance(data_type_uri, URIRef):
+                        data_type_uri = self.overwrite_data_types.get(data_type_uri, data_type_uri)
 
                     if value_type_uri := (data_type_uri or object_type_uri):
                         value_type_id = remove_namespace_from_uri(value_type_uri)