from typing import List, Optional
import uuid
from pydantic import Field
from rdflib import URIRef
from atomrdf.datamodels.basemodels import BaseModel, TemplateMixin
from atomrdf.namespace import DCAT, DCTERMS, FOAF, Literal
[docs]
class Creator(BaseModel, TemplateMixin):
"""A person who created a dataset, mapped to foaf:Person.
``id`` (from TemplateMixin): URI identifying the person, e.g. an ORCID.
"""
name: Optional[str] = Field(default=None, description="Full name (foaf:name)")
def to_graph(self, graph) -> URIRef:
uri = self.id if self.id else f"person:{uuid.uuid4()}"
subject = graph.create_node(uri, FOAF.Person, label=self.name)
if self.name:
graph.add((subject, FOAF.name, Literal(self.name)))
return subject
@classmethod
def from_graph(cls, graph, subject: URIRef) -> "Creator":
name_node = graph.value(subject, FOAF.name)
return cls(
id=str(subject),
name=str(name_node) if name_node is not None else None,
)
[docs]
class Publication(BaseModel, TemplateMixin):
"""A bibliographic resource (paper), mapped to dcterms:BibliographicResource.
``id`` (from TemplateMixin): URI identifying the paper.
"""
identifier: Optional[str] = Field(default=None, description="DOI string")
title: Optional[str] = Field(default=None, description="dcterms:title")
def to_graph(self, graph) -> URIRef:
uri = self.id if self.id else f"publication:{uuid.uuid4()}"
subject = graph.create_node(
uri, DCTERMS.BibliographicResource, label=self.title
)
if self.identifier:
graph.add((subject, DCTERMS.identifier, Literal(self.identifier)))
if self.title:
graph.add((subject, DCTERMS.title, Literal(self.title)))
return subject
@classmethod
def from_graph(cls, graph, subject: URIRef) -> "Publication":
identifier_node = graph.value(subject, DCTERMS.identifier)
title_node = graph.value(subject, DCTERMS.title)
return cls(
id=str(subject),
identifier=str(identifier_node) if identifier_node is not None else None,
title=str(title_node) if title_node is not None else None,
)
[docs]
class Dataset(BaseModel, TemplateMixin):
"""A dataset, mapped to dcat:Dataset."""
identifier: Optional[str] = Field(
default=None, description="DOI URI for the dataset"
)
title: Optional[str] = Field(default=None, description="dcterms:title")
creators: List[Creator] = Field(default_factory=list)
publication: Optional[Publication] = Field(default=None)
samples: List[str] = Field(
default_factory=list, description="Sample URIs linked via dcterms:isPartOf"
)
def to_graph(self, graph) -> URIRef:
uri = self.identifier if self.identifier else f"dataset:{uuid.uuid4()}"
subject = graph.create_node(uri, DCAT.Dataset, label=self.title)
if self.identifier:
graph.add((subject, DCTERMS.identifier, Literal(self.identifier)))
if self.title:
graph.add((subject, DCTERMS.title, Literal(self.title)))
for creator in self.creators:
person_uri = creator.to_graph(graph)
graph.add((subject, DCTERMS.creator, person_uri))
if self.publication is not None:
pub_uri = self.publication.to_graph(graph)
graph.add((subject, DCTERMS.isReferencedBy, pub_uri))
for sample_uri in self.samples:
graph.add((URIRef(sample_uri), DCTERMS.isPartOf, subject))
return subject
[docs]
@classmethod
def from_dict(cls, data: dict) -> "Dataset":
"""Construct a Dataset from a plain dictionary (e.g., parsed YAML)."""
creators = [Creator(**c) for c in data.get("creators", [])]
pub_data = data.get("publication")
publication = Publication(**pub_data) if pub_data else None
return cls(
identifier=data.get("identifier"),
title=data.get("title"),
creators=creators,
publication=publication,
samples=data.get("samples", []),
)
