Skip to content

Commit

Permalink
feat(utils): implement new type rdf importer
Browse files Browse the repository at this point in the history
  • Loading branch information
b1rger committed Jan 27, 2025
1 parent 6e9b306 commit 41d410f
Show file tree
Hide file tree
Showing 2 changed files with 74 additions and 0 deletions.
13 changes: 13 additions & 0 deletions apis_core/apis_entities/triple_configs/E21_PersonFromDNB.toml
Original file line number Diff line number Diff line change
@@ -0,0 +1,13 @@
[[filters]]
"rdf:type" = "gndo:DifferentiatedPerson"

[attributes]
forename = ["gndo:forename", "gndo:preferredNameEntityForThePerson/gndo:forename"]
alternative_names = "gndo:variantNameForThePerson"
surname = ["gndo:surname", "gndo:preferredNameEntityForThePerson/gndo:surname"]
start_date_written = "gndo:dateOfBirth"
end_date_written = "gndo:dateOfDeath"
same_as = "owl:sameAs"
profession = "gndo:professionOrOccupation"

relations = ["gndo:placeOfDeath", "gndo:placeOfBirth"]
61 changes: 61 additions & 0 deletions apis_core/utils/rdf2.py
Original file line number Diff line number Diff line change
@@ -0,0 +1,61 @@
# SPDX-FileCopyrightText: 2025 Birger Schacht
# SPDX-License-Identifier: MIT

import logging
from collections import defaultdict

from AcdhArcheAssets.uri_norm_rules import get_normalized_uri
from rdflib import BNode, Graph, RDF

from apis_core.utils.settings import dict_from_toml_directory

logger = logging.getLogger(__name__)


def find_matching_config(graph: Graph) -> dict | None:
triple_configs = dict_from_toml_directory("triple_configs")
for path, config in triple_configs.items():
for _filter in config.get("filters", []):
triples = [
(
None,
graph.namespace_manager.expand_curie(predicate),
graph.namespace_manager.expand_curie(obj),
)
for predicate, obj in _filter.items()
]
triples = [triple in graph for triple in triples]
if all(triples):
logger.debug("Using %s for parsing graph", path)
return config
return None


def get_something_from_uri(uri: str) -> dict | None:
uri = get_normalized_uri(uri)
graph = Graph()
graph.parse(uri)

if config := find_matching_config(graph):
result = defaultdict(list)
result["relations"] = defaultdict(list)

for attribute, curies in config.get("attributes", {}).items():
if isinstance(curies, str):
curies = [curies]
for curie in curies:
values = []
results = graph.query("SELECT ?object WHERE { ?subject " + curie + " ?object }")
objects = [result.object for result in results]
for obj in objects:
if isinstance(obj, BNode):
values.extend([value.toPython() for value in graph.objects(subject=obj) if value != RDF.Seq])
else:
values.append(obj.toPython())

if attribute == "relations":
result["relations"][curie].extend(values)
else:
result[attribute].extend(values)
return dict(result)
return None

0 comments on commit 41d410f

Please sign in to comment.