Merge pull request DSpace#9690 from uniba-ub/issue-9660

exclude submitter information in solr indexing

dspace-api/src/main/java/org/dspace/discovery/indexobject/ItemIndexFactoryImpl.java

@@ -154,9 +154,11 @@ public SolrInputDocument buildDocument(Context context, IndexableItem indexableI
         doc.addField("latestVersion", isLatestVersion(context, item));
 
         EPerson submitter = item.getSubmitter();
-        if (submitter != null) {
-            addFacetIndex(doc, "submitter", submitter.getID().toString(),
-                    submitter.getFullName());
+        if (submitter != null && !(DSpaceServicesFactory.getInstance().getConfigurationService().getBooleanProperty(
+            "discovery.index.item.submitter.enabled", false))) {
+            doc.addField("submitter_authority", submitter.getID().toString());
+        } else if (submitter != null) {
+            addFacetIndex(doc, "submitter", submitter.getID().toString(), submitter.getFullName());
         }
 
         // Add the item metadata

dspace-api/src/test/java/org/dspace/discovery/DiscoveryIT.java

@@ -8,6 +8,10 @@
 package org.dspace.discovery;
 
 import static org.dspace.discovery.SolrServiceWorkspaceWorkflowRestrictionPlugin.DISCOVER_WORKSPACE_CONFIGURATION_NAME;
+import static org.hamcrest.MatcherAssert.assertThat;
+import static org.hamcrest.Matchers.hasItem;
+import static org.hamcrest.Matchers.hasItems;
+import static org.hamcrest.Matchers.not;
 import static org.junit.Assert.assertEquals;
 import static org.junit.Assert.assertFalse;
 import static org.junit.Assert.assertTrue;
@@ -21,6 +25,10 @@
 import java.util.stream.Collectors;
 
 import jakarta.servlet.http.HttpServletRequest;
+import org.apache.solr.client.solrj.SolrQuery;
+import org.apache.solr.client.solrj.SolrServerException;
+import org.apache.solr.client.solrj.response.QueryResponse;
+import org.apache.solr.common.SolrDocument;
 import org.dspace.AbstractIntegrationTestWithDatabase;
 import org.dspace.app.launcher.ScriptLauncher;
 import org.dspace.app.scripts.handler.impl.TestDSpaceRunnableHandler;
@@ -99,6 +107,9 @@ public class DiscoveryIT extends AbstractIntegrationTestWithDatabase {
     MetadataAuthorityService metadataAuthorityService = ContentAuthorityServiceFactory.getInstance()
                                                                                       .getMetadataAuthorityService();
 
+    MockSolrSearchCore solrSearchCore = DSpaceServicesFactory.getInstance().getServiceManager()
+        .getServiceByName(null, MockSolrSearchCore.class);
+
     @Override
     @Before
     public void setUp() throws Exception {
@@ -796,6 +807,104 @@ public void searchWithDefaultSortServiceTest() throws SearchServiceException {
         }
     }
 
+    /**
+     * Test designed to check if the submitter is not indexed in all in solr documents for items
+     * and the submitter authority is still indexed
+     * @throws SearchServiceException
+     */
+    @Test
+    public void searchWithNoSubmitterTest() throws SearchServiceException {
+
+        configurationService.setProperty("discovery.index.item.submitter.enabled", false);
+        DiscoveryConfiguration defaultConf = SearchUtils.getDiscoveryConfiguration(context, "default", null);
+
+        // Populate the testing objects: create items in eperson's workspace and perform search in it
+        int numberItems = 10;
+        context.turnOffAuthorisationSystem();
+        EPerson submitter = null;
+        try {
+            submitter = EPersonBuilder.createEPerson(context).withEmail("[email protected]")
+                .withNameInMetadata("Peter", "Funny").build();
+        } catch (SQLException e) {
+            throw new RuntimeException(e);
+        }
+        context.setCurrentUser(submitter);
+        Community community = CommunityBuilder.createCommunity(context).build();
+        Collection collection = CollectionBuilder.createCollection(context, community).build();
+        for (int i = 0; i < numberItems; i++) {
+            ItemBuilder.createItem(context, collection)
+                .withTitle("item " + i)
+                .build();
+        }
+        context.restoreAuthSystemState();
+
+        // Build query with default parameters (except for workspaceConf)
+        QueryResponse result = null;
+        try {
+            result = solrSearchCore.getSolr().query(new SolrQuery(String.format(
+                "search.resourcetype:\"Item\"")));
+        } catch (SolrServerException e) {
+            throw new RuntimeException(e);
+        } catch (IOException e) {
+            throw new RuntimeException(e);
+        }
+        assertEquals(result.getResults().size(), numberItems);
+        for (SolrDocument doc : result.getResults()) {
+            assertThat(doc.getFieldNames(),
+                not(hasItems("submitter_keyword", "submitter_ac", "submitter_acid", "submitter_filter")));
+            assertThat(doc.getFieldNames(), hasItem("submitter_authority"));
+        }
+    }
+
+    /**
+     * Test designed to check if the submitter is indexed in all in solr documents for items
+     * @throws SearchServiceException
+     */
+    @Test
+    public void searchWithSubmitterTest() throws SearchServiceException {
+
+        configurationService.setProperty("discovery.index.item.submitter.enabled", true);
+        DiscoveryConfiguration defaultConf = SearchUtils.getDiscoveryConfiguration(context, "default", null);
+
+        // Populate the testing objects: create items in eperson's workspace and perform search in it
+        int numberItems = 10;
+        context.turnOffAuthorisationSystem();
+        EPerson submitter = null;
+        try {
+            submitter = EPersonBuilder.createEPerson(context).withEmail("[email protected]")
+                .withNameInMetadata("Peter", "Funny").build();
+        } catch (SQLException e) {
+            throw new RuntimeException(e);
+        }
+        context.setCurrentUser(submitter);
+        Community community = CommunityBuilder.createCommunity(context).build();
+        Collection collection = CollectionBuilder.createCollection(context, community).build();
+        for (int i = 0; i < numberItems; i++) {
+            ItemBuilder.createItem(context, collection)
+                .withTitle("item " + i)
+                .build();
+        }
+        context.restoreAuthSystemState();
+
+        // Build query with default parameters (except for workspaceConf)
+        QueryResponse result = null;
+        try {
+            result = solrSearchCore.getSolr().query(new SolrQuery(String.format(
+                "search.resourcetype:\"Item\"")));
+        } catch (SolrServerException e) {
+            throw new RuntimeException(e);
+        } catch (IOException e) {
+            throw new RuntimeException(e);
+        }
+        assertEquals(result.getResults().size(), numberItems);
+        for (SolrDocument doc : result.getResults()) {
+            for (String fieldname : doc.getFieldNames()) {
+                assertThat(doc.getFieldNames(), hasItems("submitter_keyword","submitter_ac", "submitter_filter",
+                    "submitter_authority"));
+            }
+        }
+    }
+
     private void assertSearchQuery(String resourceType, int size) throws SearchServiceException {
         assertSearchQuery(resourceType, size, size, 0, -1);
     }

dspace/config/modules/discovery.cfg

@@ -24,6 +24,11 @@ discovery.search.server = ${solr.server}/${solr.multicorePrefix}search
 # discovery.index.ignore-authority = false
 discovery.index.projection=dc.title,dc.contributor.*,dc.date.issued
 
+# Restricts the indexing of the submitter for archived items
+# By default the submitter information from the corresponding eperson is not indexed.
+# If you set this value to true, than the submitter information is indexed and you will need to reindex search core
+# discovery.index.item.submitter.enabled = false
+
 # Allow auto-reindexing.
 # If any database migrations are applied to your database (via Flyway), then a
 # reindex flag is always written to '[dspace]/solr/search/conf/reindex.flag'.
@@ -48,4 +53,4 @@ discovery.facet.namedtype.workflow.pooled = 004workflow\n|||\nWaiting for Contro
 # Set the number of retry of a query when stale objects are found.
 # Set to -1 if stale objects should be ignored. Set to 0 if you want to avoid extra query but take the chance to cleanup 
 # the index each time that stale objects are found. Default 3
-discovery.removestale.attempts = 3
+discovery.removestale.attempts = 3