Add function and CLI command for serving SSSOM from SPARQL (#619)

Author: cthoyt

poetry.lock

@@ -44,6 +44,12 @@ docs = [
     "sphinx-click>=4.3.0",
     "myst-parser>=0.18.1",
 ]
+rdflib-endpoint = [
+    "uvicorn",
+    "fastapi",
+    "rdflib-endpoint",
+    "httpx",  # only used in tests, but this is small
+]
 
 [project.scripts]
 sssom = "sssom.cli:main"

pyproject.toml

@@ -62,7 +62,7 @@
     sort_df_rows_columns,
     to_mapping_set_dataframe,
 )
-from .writers import WRITER_FUNCTIONS, write_table
+from .writers import WRITER_FUNCTIONS, get_rdflib_endpoint_app, write_table
 
 logging = _logging.getLogger(__name__)
 
@@ -816,5 +816,25 @@ def invert(
     write_table(msdf, output)
 
 
+@main.command(name="serve-rdf")
+@input_argument
+@click.option(
+    "--hydrate/--no-hydrate",
+    is_flag=True,
+    default=True,
+    show_default=True,
+    help="Infer S-P-O simple triples from axioms. On by default.",
+)
+@click.option("--host", default="127.0.0.1", show_default=True)
+@click.option("--port", type=int, default=8000, show_default=True)
+def serve_rdf(input: str, host: str, port: int, hydrate: bool) -> None:
+    """Serve the SSSOM file as an RDF SPARQL endpoint."""
+    import uvicorn
+
+    msdf = parse_sssom_table(input)
+    app = get_rdflib_endpoint_app(msdf, hydrate=hydrate)
+    uvicorn.run(app, host=host, port=port)
+
+
 if __name__ == "__main__":
     main()

src/sssom/cli.py

@@ -1,10 +1,13 @@
 """Serialization functions for SSSOM."""
 
+from __future__ import annotations
+
 import json
 import logging as _logging
 from contextlib import contextmanager
 from pathlib import Path
 from typing import (
+    TYPE_CHECKING,
     Any,
     Callable,
     Collection,
@@ -45,6 +48,9 @@
     sort_df_rows_columns,
 )
 
+if TYPE_CHECKING:
+    import rdflib_endpoint
+
 logging = _logging.getLogger(__name__)
 
 # noinspection PyProtectedMember
@@ -362,6 +368,45 @@ def to_rdf_graph(msdf: MappingSetDataFrame, *, hydrate: bool = False) -> Graph:
     return cast(Graph, graph)
 
 
+EXAMPLE_SPARQL_QUERY = """\
+    PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
+    PREFIX owl: <http://www.w3.org/2002/07/owl#>
+    PREFIX sssom: <https://w3id.org/sssom/>
+    PREFIX obo: <http://purl.obolibrary.org/obo/>
+    PREFIX semapv: <https://w3id.org/semapv/vocab/>
+    PREFIX skos: <http://www.w3.org/2004/02/skos/core#>
+    PREFIX pav: <http://purl.org/pav/>
+    PREFIX orcid: <https://orcid.org/>
+    PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
+
+    SELECT ?s ?p ?o ?justification {
+      [] a owl:Axiom ;
+        owl:annotatedSource ?s ;
+        owl:annotatedProperty ?p ;
+        owl:annotatedTarget ?o ;
+        sssom:mapping_justification ?justification ;
+    }
+    LIMIT 50
+"""
+
+
+def get_rdflib_endpoint_app(
+    msdf: MappingSetDataFrame, *, hydrate: bool = True
+) -> rdflib_endpoint.SparqlEndpoint:
+    """Get a FastAPI app that serves the mappings from a SPARQL endpoint."""
+    from rdflib_endpoint import SparqlEndpoint
+
+    graph = to_rdf_graph(msdf, hydrate=hydrate)
+    app = SparqlEndpoint(
+        graph=graph,
+        cors_enabled=True,
+        title=f"SSSOM SPARQL Endpoint for {msdf.metadata['mapping_set_id']}",
+        description=msdf.metadata.get("mapping_set_description"),
+        example_query=EXAMPLE_SPARQL_QUERY,
+    )
+    return app
+
+
 def to_fhir_json(msdf: MappingSetDataFrame) -> Dict[str, Any]:
     """Convert a mapping set dataframe to a JSON object.