Merge pull request #35 from ternaustralia/edmond/ld-viewer

Streamline API error responses, Improve performance and improve the fetching of labels of resources with deprecated code removed

src/linkeddata_api/domain/label.py

@@ -12,16 +12,30 @@ def get(
     """
     Returns a label or None if no label found.
     """
-    query = f"""
+    template = Template(
+        """
         PREFIX skos: <http://www.w3.org/2004/02/skos/core#>
+        PREFIX dcterms: <http://purl.org/dc/terms/>
+        PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
+        PREFIX schema: <https://schema.org/>
+        PREFIX sdo: <http://schema.org/>
+
         SELECT DISTINCT ?label
-        WHERE {{
-            VALUES (?labelProperty) {{
+        WHERE {
+            VALUES (?labelProperty) {
                 (skos:prefLabel)
-            }}
-            <{uri}> ?labelProperty ?label .
-        }}
+                (rdfs:label)
+                (dcterms:title)
+                (schema:name)
+                (sdo:name)
+                (dcterms:identifier)
+            }
+            <{{ uri }}> ?labelProperty ?label .
+        }
+        LIMIT 1
     """
+    )
+    query = template.render(uri=uri)
 
     result = data.sparql.post(query, sparql_endpoint).json()
 
@@ -41,28 +55,37 @@ def _get_from_list_query(uris: list[str]) -> str:
     template = Template(
         """
         PREFIX skos: <http://www.w3.org/2004/02/skos/core#>
-        SELECT DISTINCT ?uri (SAMPLE(?_label) AS ?label)
+        PREFIX dcterms: <http://purl.org/dc/terms/>
+        PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
+        PREFIX schema: <https://schema.org/>
+        PREFIX sdo: <http://schema.org/>
+
+        SELECT DISTINCT ?uri ?label
         WHERE {
-            VALUES (?uri) {
-                {% for uri in uris %}
-                (<{{ uri }}>)
-                {% endfor %}
-            }
-            
+            {% for uri in uris %}
             {
-                ?uri skos:prefLabel ?_label .        
-            }
-            UNION {
-                # Also try and fetch label from TERN's controlled vocabularies.
-                SERVICE <https://graphdb.tern.org.au/repositories/tern_vocabs_core> {
-                    ?uri skos:prefLabel ?_label .
+                SELECT DISTINCT ?uri ?label
+                WHERE {
+                    BIND(<{{ uri }}> as ?uri)
+                    VALUES (?labelProperty) {
+                        (skos:prefLabel)
+                        (rdfs:label)
+                        (dcterms:title)
+                        (schema:name)
+                        (sdo:name)
+                        (dcterms:identifier)
+                    }
+                    ?uri ?labelProperty ?label .
                 }
+                LIMIT 1
             }
+            {% if not loop.last %}UNION{% endif %}
+            {% endfor %}
         }
-        GROUP BY ?uri
     """
     )
-    return template.render(uris=uris)
+    query = template.render(uris=uris)
+    return query
 
 
 def get_from_list(

src/linkeddata_api/domain/viewer/resource/json/__init__.py

@@ -244,16 +244,21 @@ def _get_types_and_properties(
 
     for row in result["results"]["bindings"]:
         if row["p"]["value"] == str(RDF.type):
-            type_label = uri_label_index.get(row["o"]["value"]) or domain.curie.get(
-                row["o"]["value"]
-            )
-            types.append(
-                domain.schema.URI(
-                    label=type_label,
-                    value=row["o"]["value"],
-                    internal=uri_internal_index.get(row["o"]["value"], False),
+            if row["o"]["type"] != "bnode":
+                type_label = uri_label_index.get(row["o"]["value"]) or domain.curie.get(
+                    row["o"]["value"]
                 )
-            )
+                types.append(
+                    domain.schema.URI(
+                        label=type_label,
+                        value=row["o"]["value"],
+                        internal=uri_internal_index.get(row["o"]["value"], False),
+                    )
+                )
+            else:
+                # TODO: handle types that have bnode values
+                # E.g. /viewers/general?uri=http://linked.data.gov.au/dataset/ausplots/site-ntabrt0001&sparql_endpoint=https://graphdb.tern.org.au/repositories/knowledge_graph_core
+                continue
         else:
             predicate_label = domain.curie.get(row["p"]["value"])
             predicate = domain.schema.URI(
@@ -317,7 +322,8 @@ def _get_types_and_properties(
             for p in properties:
                 if p.predicate.value == predicate.value:
                     found = True
-                    p.objects.append(item)
+                    if item not in p.objects:
+                        p.objects.append(item)
 
             if not found:
                 properties.append(

src/linkeddata_api/views/api_v1/__init__.py

@@ -2,7 +2,6 @@
 
 # import all sub modules with views registered with blueprint
 from . import ontology_viewer
-from . import vocab_viewer
 from . import version_info
 from . import rdf_tools
 from . import viewer

src/linkeddata_api/views/api_v1/openapi.yaml

@@ -56,10 +56,14 @@ components:
             type: string
           created:
             title: The date when the resource was created
-            type: string
+            oneOf:
+              - type: string
+            nullable: true
           modified: 
             title: The date when the resource was modified
-            type: string
+            oneOf:
+              - type: string
+            nullable: true
     Resource:
       title: Resource
       type: object
@@ -174,6 +178,8 @@ paths:
             application/json:
               schema:
                 $ref: "#/components/schemas/EntrypointItemList"
+        "404":
+          description: The supplied `viewer_id` value was not found
         "502":
           description: Gateway error
   /viewer/resource:
@@ -246,24 +252,14 @@ paths:
             application/json:
               schema:
                 $ref: "#/components/schemas/Resource"
+        "400":
+          description: Client error
         "404":
           description: Resource of URI not found.
-          content:
-            text/html:
-              schema:
-                type: string
         "500":
           description: Internal server error
-          content:
-            text/plain:
-              schema:
-                type: string
         "502":
           description: Error communicating with the database.
-          content:
-            text/html:
-              schema:
-                type: string
 
   /ontology_viewer/classes/flat:
     get:
@@ -302,95 +298,6 @@ paths:
                 type: string
         "502":
           description: Gateway error
-
-  /vocab_viewer/nrm/vocabs:
-    get:
-      tags:
-        - NRM vocabularies
-      summary: Get a list of NRM protocol vocabularies
-      description: Get a list of concept schemes and collections for the NRM protocol.
-      responses:
-        "200":
-          description: A list of concept schemes and collections.
-          content:
-            application/json:
-              schema:
-                type: array
-                items:
-                  title: Vocabulary item
-                  type: object
-                  properties:
-                    id:
-                      title: IRI of item
-                      type: string
-                    label:
-                      title: Label of item
-                      type: string
-                    description:
-                      title: The description of the item
-                      type: string
-                    created:
-                      title: The date when the resource was created
-                      type: string
-                    modified: 
-                      title: The date when the resource was modified
-                      type: string
-        "502":
-          description: Gateway error
-
-  /vocab_viewer/nrm/resource:
-    get:
-      tags: 
-        - NRM vocabularies
-      summary: Get a resource's description
-      description: A JSON payload used to render frontend user-interfaces.
-      parameters:
-        - in: query
-          name: uri
-          schema:
-            type: string
-          required: true
-          description: URI of the resource.
-          examples:
-            nrm:
-              summary: NRM index
-              value: https://linked.data.gov.au/def/nrm
-        - in: query
-          name: sparql_endpoint
-          schema:
-            type: string
-          required: true
-          description: SPARQL endpoint
-          examples:
-            nrm:
-              summary: NRM SPARQL endpoint
-              value: https://graphdb.tern.org.au/repositories/dawe_vocabs_core
-      responses:
-        "200":
-          description: A resource's description
-          content:
-            application/json:
-              schema:
-                type: object
-                properties:
-                  uri:
-                    type: string
-                  label:
-                    type: string
-                  types:
-                    type: array
-                    items:
-                      $ref: "#/components/schemas/URI"
-                  properties:
-                    type: array
-                    items:
-                      $ref: "#/components/schemas/PredicateObjects"
-        "404":
-          description: Resource with URI not found
-          content:
-            text/plain:
-              schema:
-                type: string
 
   /rdf_tools/convert:
     post:

src/linkeddata_api/views/api_v1/viewer/entrypoint.py

@@ -1,5 +1,4 @@
-from flask import Response
-from werkzeug.exceptions import HTTPException
+from flask import abort
 from flask_tern import openapi
 
 from linkeddata_api.views.api_v1.blueprint import bp
@@ -31,10 +30,10 @@ def get_entrypoint(viewer_id: str):
         sparql_endpoint = obj["sparql_endpoint"]
         items = obj["func"](sparql_endpoint)
     except ViewerIDNotFoundError as err:
-        raise HTTPException(str(err), Response(str(err), 404)) from err
+        abort(404, str(err))
     except (RequestError, SPARQLResultJSONError) as err:
-        raise HTTPException(err.description, Response(err.description, 502)) from err
+        abort(502, err.description)
     except Exception as err:
-        raise HTTPException(str(err), Response(str(err), 500)) from err
+        abort(500, err)
 
     return jsonify(items, headers={"cache-control": "max-age=600, s-maxage=3600"})

src/linkeddata_api/views/api_v1/viewer/resource.py

@@ -1,7 +1,6 @@
 import logging
 
-from flask import request, Response
-from werkzeug.exceptions import HTTPException
+from flask import request, Response, abort
 from flask_tern import openapi
 
 from linkeddata_api.views.api_v1.blueprint import bp
@@ -29,9 +28,11 @@ def get_resource():
         True if include_incoming_relationships == "true" else False
     )
 
-    if uri is None or sparql_endpoint is None:
-        err_msg = "Required query parameters 'uri' or 'sparql_endpoint' was not provided."
-        raise HTTPException(err_msg, Response(err_msg, 404))
+    if not uri or not sparql_endpoint:
+        err_msg = (
+            "Required query parameters 'uri' or 'sparql_endpoint' was not provided."
+        )
+        abort(400, err_msg)
 
     logger.info(
         """
@@ -57,17 +58,11 @@ def get_resource():
             uri, sparql_endpoint, format_, include_incoming_relationships
         )
     except SPARQLNotFoundError as err:
-        raise HTTPException(err.description, Response(err.description, 404)) from err
+        abort(404, err.description)
     except (RequestError, SPARQLResultJSONError) as err:
-        raise HTTPException(
-            description=err.description,
-            response=Response(err.description, status=502),
-        ) from err
+        abort(502, err.description)
     except Exception as err:
-        raise HTTPException(
-            description=str(err),
-            response=Response(str(err), mimetype="text/plain", status=500),
-        ) from err
+        abort(500, err)
 
     return Response(
         result,