feat(api-v2): Allow querying for rdfs:label in Gravsearch (#1649)

* feat(api-v2): Make the inferring type inspector know the object type of rdfs:label. * feat(api-v2): Support Gravsearch queries using rdfs:label (ongoing). * feat(api-v2): Support wildcard searches in resource labels. * feat(api-v2): Support Lucene searches in rdfs:label (ongoing). * test(gravsearch): Add tests for wildcard searches with rdfs:label. * feat(gravsearch): Support the regex function with rdfs:label. - Add tests. - Add docs. * test(gravsearch): Fix tests. * style(gravsearch): Use apply() instead of newInstance().
dasch-swiss · Jun 3, 2020 · d56004b · d56004b
1 parent bd08f2c
commit d56004b
Show file tree

Hide file tree

Showing 12 changed files with 1,579 additions and 672 deletions.
diff --git a/docs/src/paradox/03-apis/api-v2/query-language.md b/docs/src/paradox/03-apis/api-v2/query-language.md
@@ -616,6 +616,39 @@ CONSTRUCT {
 }
 ```
 
+### Filtering on `rdfs:label`
+
+The `rdfs:label` of a resource is not a Knora value, but you can still search for it.
+This can be done in the same ways in the simple or complex schema:
+
+Using a string literal object:
+
+```
+?book rdfs:label "Zeitglöcklein des Lebens und Leidens Christi" .
+```
+
+Using a variable and a FILTER:
+
+```
+?book rdfs:label ?label .
+FILTER(?label = "Zeitglöcklein des Lebens und Leidens Christi")
+```
+
+Using the `regex` function:
+
+```
+?book rdfs:label ?bookLabel .
+FILTER regex(?bookLabel, "Zeit", "i")
+```
+
+To match words in an `rdfs:label` using the full-text search index, use the
+`knora-api:matchLabel` function, which works like `knora-api:matchText`,
+except that the first argument is a variable representing a resource:
+
+```
+FILTER knora-api:matchLabel(?book, "Zeitglöcklein")
+```
+
 ### CONSTRUCT Clause
 
 In the `CONSTRUCT` clause of a Gravsearch query, the variable representing the

diff --git a/webapi/src/main/scala/org/knora/webapi/OntologyConstants.scala b/webapi/src/main/scala/org/knora/webapi/OntologyConstants.scala
@@ -133,6 +133,13 @@ object OntologyConstants {
         val InternalOntologyStart = "http://www.knora.org/ontology"
     }
 
+    /**
+     * The object types of resource metadata properties.
+     */
+    val ResourceMetadataPropertyAxioms: Map[IRI, IRI] = Map(
+        OntologyConstants.Rdfs.Label -> OntologyConstants.Xsd.String
+    )
+
     /**
      * Ontology labels that are reserved for built-in ontologies.
      */
@@ -922,6 +929,7 @@ object OntologyConstants {
         val MatchTextFunction: IRI = KnoraApiV2PrefixExpansion + "matchText"
         val MatchInStandoffFunction: IRI = KnoraApiV2PrefixExpansion + "matchInStandoff"
         val MatchTextInStandoffFunction: IRI = KnoraApiV2PrefixExpansion + "matchTextInStandoff"
+        val MatchLabelFunction: IRI = KnoraApiV2PrefixExpansion + "matchLabel"
         val StandoffLinkFunction: IRI = KnoraApiV2PrefixExpansion + "standoffLink"
     }
 
@@ -961,6 +969,7 @@ object OntologyConstants {
         val MatchesTextIndex: IRI = KnoraApiV2PrefixExpansion + "matchesTextIndex" // virtual property to be replaced by a triplestore-specific one
         val MatchFunction: IRI = KnoraApiV2PrefixExpansion + "match"
         val MatchTextFunction: IRI = KnoraApiV2PrefixExpansion + "matchText"
+        val MatchLabelFunction: IRI = KnoraApiV2PrefixExpansion + "matchLabel"
 
         val ResourceProperty: IRI = KnoraApiV2PrefixExpansion + "resourceProperty"
 
@@ -1135,5 +1144,4 @@ object OntologyConstants {
         val KnoraExplicitNamedGraph: IRI = "http://www.knora.org/explicit"
         val GraphDBExplicitNamedGraph: IRI = "http://www.ontotext.com/explicit"
     }
-
 }
diff --git a/.../main/scala/org/knora/webapi/responders/v2/search/gravsearch/GravsearchQueryChecker.scala b/.../main/scala/org/knora/webapi/responders/v2/search/gravsearch/GravsearchQueryChecker.scala
@@ -135,7 +135,6 @@ object GravsearchQueryChecker {
 
     // A set of predicates that aren't allowed in Gravsearch.
     val forbiddenPredicates: Set[IRI] = Set(
-        OntologyConstants.Rdfs.Label,
         OntologyConstants.KnoraApiV2Complex.AttachedToUser,
         OntologyConstants.KnoraApiV2Complex.HasPermissions,
         OntologyConstants.KnoraApiV2Complex.CreationDate,

diff --git a/...org/knora/webapi/responders/v2/search/gravsearch/prequery/AbstractPrequeryGenerator.scala b/...org/knora/webapi/responders/v2/search/gravsearch/prequery/AbstractPrequeryGenerator.scala
diff --git a/...g/knora/webapi/responders/v2/search/gravsearch/types/GravsearchTypeInspectionRunner.scala b/...g/knora/webapi/responders/v2/search/gravsearch/types/GravsearchTypeInspectionRunner.scala
@@ -22,16 +22,17 @@ package org.knora.webapi.responders.v2.search.gravsearch.types
 import org.knora.webapi.messages.admin.responder.usersmessages.UserADM
 import org.knora.webapi.responders.ResponderData
 import org.knora.webapi.responders.v2.search._
+import org.knora.webapi.util.StringFormatter
 import org.knora.webapi.{GravsearchException, KnoraDispatchers, OntologyConstants}
 
 import scala.concurrent.{ExecutionContext, Future}
 
 /**
-  * Runs Gravsearch type inspection using one or more type inspector implementations.
-  *
-  * @param system     the Akka actor system.
-  * @param inferTypes if true, use type inference.
-  */
+ * Runs Gravsearch type inspection using one or more type inspector implementations.
+ *
+ * @param responderData the Knora [[ResponderData]].
+ * @param inferTypes    if true, use type inference.
+ */
 class GravsearchTypeInspectionRunner(responderData: ResponderData,
                                      inferTypes: Boolean = true) {
     private implicit val executionContext: ExecutionContext = responderData.system.dispatchers.lookup(KnoraDispatchers.KnoraActorDispatcher)
@@ -55,15 +56,17 @@ class GravsearchTypeInspectionRunner(responderData: ResponderData,
     )
 
     /**
-      * Given the WHERE clause from a parsed Gravsearch query, returns information about the types found
-      * in the query.
-      *
-      * @param whereClause    the Gravsearch WHERE clause.
-      * @param requestingUser the requesting user.
-      * @return the result of the type inspection.
-      */
+     * Given the WHERE clause from a parsed Gravsearch query, returns information about the types found
+     * in the query.
+     *
+     * @param whereClause    the Gravsearch WHERE clause.
+     * @param requestingUser the requesting user.
+     * @return the result of the type inspection.
+     */
     def inspectTypes(whereClause: WhereClause,
                      requestingUser: UserADM): Future[GravsearchTypeInspectionResult] = {
+        implicit val stringFormatter: StringFormatter = StringFormatter.getGeneralInstance
+
         for {
             // Get the set of typeable entities in the Gravsearch query.
             typeableEntities: Set[TypeableEntity] <- Future {
@@ -110,16 +113,16 @@ class GravsearchTypeInspectionRunner(responderData: ResponderData,
 
 
     /**
-      * A [[WhereVisitor]] that collects typeable entities from a Gravsearch WHERE clause.
-      */
+     * A [[WhereVisitor]] that collects typeable entities from a Gravsearch WHERE clause.
+     */
     private class TypeableEntityCollectingWhereVisitor extends WhereVisitor[Set[TypeableEntity]] {
         /**
-          * Collects typeable entities from a statement.
-          *
-          * @param statementPattern the pattern to be visited.
-          * @param acc              the accumulator.
-          * @return the accumulator.
-          */
+         * Collects typeable entities from a statement.
+         *
+         * @param statementPattern the pattern to be visited.
+         * @param acc              the accumulator.
+         * @return the accumulator.
+         */
         override def visitStatementInWhere(statementPattern: StatementPattern, acc: Set[TypeableEntity]): Set[TypeableEntity] = {
             statementPattern.pred match {
                 case iriRef: IriRef if iriRef.iri.toString == OntologyConstants.Rdf.Type =>
@@ -133,23 +136,23 @@ class GravsearchTypeInspectionRunner(responderData: ResponderData,
         }
 
         /**
-          * Collects typeable entities from a `FILTER`.
-          *
-          * @param filterPattern the pattern to be visited.
-          * @param acc           the accumulator.
-          * @return the accumulator.
-          */
+         * Collects typeable entities from a `FILTER`.
+         *
+         * @param filterPattern the pattern to be visited.
+         * @param acc           the accumulator.
+         * @return the accumulator.
+         */
         override def visitFilter(filterPattern: FilterPattern, acc: Set[TypeableEntity]): Set[TypeableEntity] = {
             visitFilterExpression(filterPattern.expression, acc)
         }
 
         /**
-          * Collects typeable entities from a filter expression.
-          *
-          * @param filterExpression the filter expression to be visited.
-          * @param acc              the accumulator.
-          * @return the accumulator.
-          */
+         * Collects typeable entities from a filter expression.
+         *
+         * @param filterExpression the filter expression to be visited.
+         * @param acc              the accumulator.
+         * @return the accumulator.
+         */
         private def visitFilterExpression(filterExpression: Expression, acc: Set[TypeableEntity]): Set[TypeableEntity] = {
             filterExpression match {
                 case compareExpr: CompareExpression =>

diff --git a/...knora/webapi/responders/v2/search/gravsearch/types/InferringGravsearchTypeInspector.scala b/...knora/webapi/responders/v2/search/gravsearch/types/InferringGravsearchTypeInspector.scala
@@ -265,7 +265,7 @@ class InferringGravsearchTypeInspector(nextInspector: Option[GravsearchTypeInspe
                                                     None
                                             }
 
-                                        case other =>
+                                        case _ =>
                                             // We don't have the predicate's type.
                                             Set.empty[GravsearchEntityTypeInfo]
                                     }
@@ -1004,6 +1004,26 @@ class InferringGravsearchTypeInspector(nextInspector: Option[GravsearchTypeInspe
                                     (textVar -> (currentTextVarTypesFromFilters + OntologyConstants.KnoraApiV2Complex.TextValue.toSmartIri))
                             )
 
+                        case OntologyConstants.KnoraApiV2Simple.MatchLabelFunction =>
+                            // The first argument is a variable representing a resource.
+                            val resourceVar = TypeableVariable(functionCallExpression.getArgAsQueryVar(0).variableName)
+                            val currentResourceVarTypesFromFilters: Set[SmartIri] = acc.typedEntitiesInFilters.getOrElse(resourceVar, Set.empty[SmartIri])
+
+                            acc.copy(
+                                typedEntitiesInFilters = acc.typedEntitiesInFilters +
+                                    (resourceVar -> (currentResourceVarTypesFromFilters + OntologyConstants.KnoraApiV2Simple.Resource.toSmartIri))
+                            )
+
+                        case OntologyConstants.KnoraApiV2Complex.MatchLabelFunction =>
+                            // The first argument is a variable representing a resource.
+                            val resourceVar = TypeableVariable(functionCallExpression.getArgAsQueryVar(0).variableName)
+                            val currentResourceVarTypesFromFilters: Set[SmartIri] = acc.typedEntitiesInFilters.getOrElse(resourceVar, Set.empty[SmartIri])
+
+                            acc.copy(
+                                typedEntitiesInFilters = acc.typedEntitiesInFilters +
+                                    (resourceVar -> (currentResourceVarTypesFromFilters + OntologyConstants.KnoraApiV2Complex.Resource.toSmartIri))
+                            )
+
                         case OntologyConstants.KnoraApiV2Complex.MatchInStandoffFunction =>
                             // The first argument is a variable representing a string.
                             val textVar = TypeableVariable(functionCallExpression.getArgAsQueryVar(0).variableName)

diff --git a/...knora/webapi/responders/v2/search/gravsearch/types/IntermediateTypeInspectionResult.scala b/...knora/webapi/responders/v2/search/gravsearch/types/IntermediateTypeInspectionResult.scala
@@ -19,7 +19,9 @@
 
 package org.knora.webapi.responders.v2.search.gravsearch.types
 
-import org.knora.webapi.AssertionException
+import org.knora.webapi.util.IriConversions._
+import org.knora.webapi.util.StringFormatter
+import org.knora.webapi.{AssertionException, IRI, OntologyConstants}
 
 /**
   * Represents an intermediate result during type inspection. This is different from [[GravsearchTypeInspectionResult]]
@@ -79,12 +81,29 @@ case class IntermediateTypeInspectionResult(entities: Map[TypeableEntity, Set[Gr
 
 object IntermediateTypeInspectionResult {
     /**
-      * Constructs an [[IntermediateTypeInspectionResult]] for the given set of typeable entities, with no
-      * types specified.
+      * Constructs an [[IntermediateTypeInspectionResult]] for the given set of typeable entities, with built-in
+      * types specified (e.g. for `rdfs:label`).
       *
       * @param entities the set of typeable entities found in the WHERE clause of a Gravsearch query.
       */
-    def apply(entities: Set[TypeableEntity]): IntermediateTypeInspectionResult = {
-        new IntermediateTypeInspectionResult(entities = entities.map(entity => entity -> Set.empty[GravsearchEntityTypeInfo]).toMap)
+    def apply(entities: Set[TypeableEntity])(implicit stringFormatter: StringFormatter): IntermediateTypeInspectionResult = {
+        // Make an IntermediateTypeInspectionResult in which each typeable entity has no types.
+        val emptyResult = new IntermediateTypeInspectionResult(entities = entities.map(entity => entity -> Set.empty[GravsearchEntityTypeInfo]).toMap)
+
+        // Collect the typeable IRIs used.
+        val irisUsed: Set[IRI] = entities.collect {
+            case typeableIri: TypeableIri => typeableIri.iri.toString
+        }
+
+        // Find the IRIs that represent resource metadata properties, and get their object types.
+        val resourceMetadataPropertyTypesUsed: Map[TypeableIri, PropertyTypeInfo] = OntologyConstants.ResourceMetadataPropertyAxioms.filterKeys(irisUsed).map {
+            case (propertyIri, objectTypeIri) => TypeableIri(propertyIri.toSmartIri) -> PropertyTypeInfo(objectTypeIri = objectTypeIri.toSmartIri)
+        }
+
+        // Add those types to the IntermediateTypeInspectionResult.
+        resourceMetadataPropertyTypesUsed.foldLeft(emptyResult) {
+            case (acc: IntermediateTypeInspectionResult, (entity: TypeableIri, entityType: PropertyTypeInfo)) =>
+                acc.addTypes(entity, Set(entityType))
+        }
     }
 }
diff --git a/webapi/src/test/resources/test-data/searchR2RV2/ZeitglöckleinViaLabel.jsonld b/webapi/src/test/resources/test-data/searchR2RV2/ZeitglöckleinViaLabel.jsonld
@@ -0,0 +1,58 @@
+{
+  "@graph" : [ {
+    "@id" : "http://rdfh.ch/0803/c5058f3a",
+    "@type" : "incunabula:book",
+    "knora-api:arkUrl" : {
+      "@type" : "xsd:anyURI",
+      "@value" : "http://0.0.0.0:3336/ark:/72163/1/0803/c5058f3a5"
+    },
+    "knora-api:attachedToProject" : {
+      "@id" : "http://rdfh.ch/projects/0803"
+    },
+    "knora-api:attachedToUser" : {
+      "@id" : "http://rdfh.ch/users/91e19f1e01"
+    },
+    "knora-api:creationDate" : {
+      "@type" : "xsd:dateTimeStamp",
+      "@value" : "2016-03-02T15:05:10Z"
+    },
+    "knora-api:hasPermissions" : "CR knora-admin:Creator|M knora-admin:ProjectMember|V knora-admin:KnownUser|RV knora-admin:UnknownUser",
+    "knora-api:userHasPermission" : "RV",
+    "knora-api:versionArkUrl" : {
+      "@type" : "xsd:anyURI",
+      "@value" : "http://0.0.0.0:3336/ark:/72163/1/0803/c5058f3a5.20160302T150510Z"
+    },
+    "rdfs:label" : "Zeitglöcklein des Lebens und Leidens Christi"
+  }, {
+    "@id" : "http://rdfh.ch/0803/ff17e5ef9601",
+    "@type" : "incunabula:book",
+    "knora-api:arkUrl" : {
+      "@type" : "xsd:anyURI",
+      "@value" : "http://0.0.0.0:3336/ark:/72163/1/0803/ff17e5ef9601j"
+    },
+    "knora-api:attachedToProject" : {
+      "@id" : "http://rdfh.ch/projects/0803"
+    },
+    "knora-api:attachedToUser" : {
+      "@id" : "http://rdfh.ch/users/91e19f1e01"
+    },
+    "knora-api:creationDate" : {
+      "@type" : "xsd:dateTimeStamp",
+      "@value" : "2016-03-02T15:05:23Z"
+    },
+    "knora-api:hasPermissions" : "CR knora-admin:Creator|M knora-admin:ProjectMember|V knora-admin:KnownUser|RV knora-admin:UnknownUser",
+    "knora-api:userHasPermission" : "RV",
+    "knora-api:versionArkUrl" : {
+      "@type" : "xsd:anyURI",
+      "@value" : "http://0.0.0.0:3336/ark:/72163/1/0803/ff17e5ef9601j.20160302T150523Z"
+    },
+    "rdfs:label" : "Zeitglöcklein des Lebens und Leidens Christi"
+  } ],
+  "@context" : {
+    "rdf" : "http://www.w3.org/1999/02/22-rdf-syntax-ns#",
+    "knora-api" : "http://api.knora.org/ontology/knora-api/v2#",
+    "rdfs" : "http://www.w3.org/2000/01/rdf-schema#",
+    "incunabula" : "http://0.0.0.0:3333/ontology/0803/incunabula/v2#",
+    "xsd" : "http://www.w3.org/2001/XMLSchema#"
+  }
+}