Added SPARQL endpoint and full-text search.

firm_server/cli/__init__.py

@@ -23,13 +23,18 @@ class Context:
 
 
 @click.group(context_settings=dict(auto_envvar_prefix="FIRM"))
-@click.option("--config", type=click.File("r"), envvar="FIRM_CONFIG")
+@click.option(
+    "--config",
+    type=click.File("r"),
+    envvar="FIRM_CONFIG",
+)
 @click.option(
     "--storage",
     "storage_key",
     type=click.Choice(STORE_DRIVERS),
     default="filesystem",
     show_default=True,
+    envvar="FIRM_STORAGE",
 )
 @click.pass_context
 def cli(ctx: click.Context, config: click.File, storage_key: str):

firm_server/routes.py

@@ -19,14 +19,17 @@
 from firm.services.activitypub import ActivityPubService, ActivityPubTenant
 from firm.services.nodeinfo import nodeinfo_index, nodeinfo_version
 from firm.services.webfinger import webfinger
+from firm_ld.search import IndexedResource, SearchEngine
+from firm_ld.sparql import create_sparql_endpoint
+from firm_ld.store import RdfResourceStore
 from starlette.exceptions import HTTPException
 from starlette.middleware import Middleware
 from starlette.middleware.authentication import AuthenticationMiddleware
 from starlette.requests import Request
 from starlette.responses import JSONResponse
 from starlette.responses import PlainTextResponse as StarlettePlainTextResponse
 from starlette.responses import Response
-from starlette.routing import Match, Route, Scope
+from starlette.routing import Match, Mount, Route, Scope
 
 from firm_server.adapters import (
     AuthenticationBackendAdapter,
@@ -206,6 +209,48 @@ def matches(self, scope: Scope) -> tuple[Match, Scope]:
         )
 
 
+def _rdf_search(store: RdfResourceStore) -> HttpResponse:
+    # TODO support named graphs for search
+    search_engine = SearchEngine(store.graph)
+    search_engine.add_index(
+        IndexedResource(
+            "https://www.w3.org/ns/activitystreams#Note",
+            [
+                "https://www.w3.org/ns/activitystreams#content",
+                "https://www.w3.org/ns/activitystreams#summary",
+            ],
+            [
+                "https://www.w3.org/ns/activitystreams#content",
+                "https://www.w3.org/ns/activitystreams#summary",
+            ],
+        )
+    )
+    search_engine.add_index(
+        IndexedResource(
+            "https://www.w3.org/ns/activitystreams#Person",
+            [
+                "https://www.w3.org/ns/activitystreams#summary",
+            ],
+            [
+                "https://www.w3.org/ns/activitystreams#summary",
+                "https://www.w3.org/ns/activitystreams#name",
+                "https://www.w3.org/ns/activitystreams#preferredUsername",
+            ],
+        )
+    )
+    log.info("Indexing RDF store")
+    search_engine.update_index()
+
+    def _search(request: HttpRequest) -> HttpResponse:
+        # TODO Need to expose query params on HttpRequest
+        return JSONResponse(
+            search_engine.search(request.query_params["q"]),
+            headers={"Access-Control-Allow-Origin", "*"},
+        )
+
+    return _search
+
+
 def get_routes(store: ResourceStore, config: ServerConfig):
     activitypub_service = ActivityPubService(
         [
@@ -240,7 +285,7 @@ def get_routes(store: ResourceStore, config: ServerConfig):
             )
         ],
     )
-    return [
+    routes = [
         Route("/.well-known/webfinger", endpoint=_adapt_endpoint(webfinger, store)),
         Route("/.well-known/nodeinfo", endpoint=_adapt_endpoint(nodeinfo_index, store)),
         Route("/nodeinfo/{version}", endpoint=_adapt_endpoint(nodeinfo_version, store)),
@@ -252,3 +297,24 @@ def get_routes(store: ResourceStore, config: ServerConfig):
         ),
         activitypub_route,
     ]
+    if isinstance(store, RdfResourceStore):
+        log.info("Registering SPARQL endpoint")
+        example_query = """\
+PREFIX as: <https://www.w3.org/ns/activitystreams#>
+PREFIX firm: <https://firm.stevebate.dev#>
+
+SELECT ?object WHERE {
+    ?object is as:Note
+}
+""".rstrip()
+        # TODO tenant-specific config for sparql endpoints
+        # The namespace will always be firm.stevebate.dev
+        sparql_app = create_sparql_endpoint(
+            "https://firm.stevebate.dev/sparql/",
+            example_query=example_query,
+            favicon="https://firm.stevebate.dev/static/favicon/favicon.ico",
+        )
+        routes.insert(4, Mount("/sparql", app=sparql_app, name="sparql"))
+        # Add a search engine
+        routes.insert(4, Route("/search", endpoint=_rdf_search(store)))
+    return routes