Skip to main content

UpdatePersonPipeline

What it is​

  • A pipeline that adds missing person-related triples (LinkedIn page, skill, first/last name, date of birth) into an RDF triple store.
  • It does not remove or overwrite existing triples; it only inserts a triple if the exact triple does not already exist for the person.

Public API​

  • UpdatePersonPipelineConfiguration(triple_store: ITripleStoreService)

    • Configuration container providing the triple store service.

  • UpdatePersonPipelineParameters

    • Pydantic-style parameter model (PipelineParameters) with fields:
      • individual_uri: str (required; must match URI_REGEX)
      • first_name: Optional[str]
      • last_name: Optional[str]
      • date_of_birth: Optional[str] (format YYYY-MM-DD)
      • linkedin_page_uri: Optional[str] (must match URI_REGEX)
      • skill_uri: str (declared as str but default is None in Field; treat as optional in practice only if your validation allows it)

  • UpdatePersonPipeline(configuration: UpdatePersonPipelineConfiguration)

    • run(parameters: PipelineParameters) -> rdflib.Graph
      • Validates parameters type is UpdatePersonPipelineParameters.
      • Fetches the existing subject graph via triple_store.get_subject_graph(individual_uri).
      • Builds an insertion graph containing only missing triples.
      • Inserts via triple_store.insert(graph_insert).
      • Returns the merged graph (existing + inserted).
    • as_tools() -> list[langchain_core.tools.BaseTool]
      • Exposes the pipeline as a LangChain StructuredTool named update_person.
    • as_api(...) -> None
      • Present but currently does nothing (returns None).

Configuration/Dependencies​

  • Requires an ITripleStoreService implementation that provides:
    • get_subject_graph(subject: URIRef) -> rdflib.Graph
    • insert(graph: rdflib.Graph) -> None
  • Uses RDF terms/predicates from naas_abi_core.utils.Graph.ABI:
    • ABI.hasLinkedInPage, ABI.hasSkill, ABI.first_name, ABI.last_name, ABI.date_of_birth
  • Uses rdflib for RDF graph manipulation and XSD.date for date datatype.

Usage​

from naas_abi.pipelines.UpdatePersonPipeline import (
    UpdatePersonPipeline,
    UpdatePersonPipelineConfiguration,
    UpdatePersonPipelineParameters,
)

# triple_store must implement ITripleStoreService
config = UpdatePersonPipelineConfiguration(triple_store=triple_store)

pipeline = UpdatePersonPipeline(config)

result_graph = pipeline.run(
    UpdatePersonPipelineParameters(
        individual_uri="https://www.commoncoreontologies.org/ont00001262/Florent_Ravenel",
        first_name="Florent",
        last_name="Ravenel",
        date_of_birth="1990-01-01",
        linkedin_page_uri="https://example.com/linkedin/profile",
        skill_uri="https://example.com/skill/python",
    )
)

print(len(result_graph))

Caveats​

  • No updates/overwrites: if a different value already exists (e.g., another first_name), this pipeline will add the new triple rather than replacing old ones.
  • as_api() is a stub and does not register any routes.
  • skill_uri is typed as str but defined with Field(None, ...); depending on your parameter validation setup, passing None may error.