VIVO-662 Remove duplicate rows from the SPARQL query results.

Two rows are considered to be duplicates if they have the same authorshipURI or the same authorURI.

webapp/src/edu/cornell/mannlib/vitro/webapp/dao/jena/QueryUtils.java

@@ -2,10 +2,14 @@
 
 package edu.cornell.mannlib.vitro.webapp.dao.jena;
 
+import java.util.ArrayList;
+import java.util.Arrays;
 import java.util.HashMap;
 import java.util.Iterator;
+import java.util.List;
 import java.util.Map;
 
+import org.apache.commons.lang.ObjectUtils;
 import org.apache.commons.lang.StringUtils;
 import org.apache.commons.logging.Log;
 import org.apache.commons.logging.LogFactory;
@@ -56,6 +60,39 @@ public class QueryUtils {
         return map;
     }
     
+	/**
+	 * If any pair of maps in the list has the same (non-null) value for any of
+	 * these keys, call the maps duplicates and keep only the first of them.
+	 */
+	public static List<Map<String, String>> removeDuplicatesMapsFromList(
+			List<Map<String, String>> rawList, String... keys) {
+		List<Map<String, String>> filteredList = new ArrayList<>();
+		outerLoop: for (Map<String, String> rawMap : rawList) {
+			for (Map<String, String> filteredMap : filteredList) {
+				for (String key : keys) {
+					String rawValue = rawMap.get(key);
+					if (rawValue != null) {
+						if (rawValue.equals(filteredMap.get(key))) {
+							if (log.isDebugEnabled()) {
+								logDuplicateRows(rawMap, filteredMap, keys);
+							}
+							continue outerLoop;
+						}
+					}
+				}
+			}
+			filteredList.add(rawMap);
+		}
+		return filteredList;
+	}
+
+	private static void logDuplicateRows(Map<String, String> rawMap,
+			Map<String, String> filteredMap, String... keys) {
+		log.debug("Found duplicate rows, by at least one of these keys: "
+				+ Arrays.toString(keys) + ". Keeping " + filteredMap
+				+ ". Discarding " + rawMap + ".");
+	}
+    
     public static Object nodeToObject( RDFNode node ){
         if( node == null ){
             return "";

webapp/test/edu/cornell/mannlib/vitro/webapp/dao/jena/QueryUtilsTest.java

@@ -3,6 +3,14 @@
 package edu.cornell.mannlib.vitro.webapp.dao.jena;
 
 import static org.junit.Assert.*;
+import static org.junit.Assert.assertEquals;
+import static org.junit.Assert.fail;
+
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.HashMap;
+import java.util.List;
+import java.util.Map;
 
 import org.junit.Ignore;
 import org.junit.Test;
@@ -18,6 +26,11 @@ import edu.cornell.mannlib.vitro.testing.AbstractTestClass;
  * TODO
  */
 public class QueryUtilsTest extends AbstractTestClass {
+
+	// ----------------------------------------------------------------------
+	// Test bindVariables
+	// ----------------------------------------------------------------------
+
 	private QuerySolutionMap bindings = new QuerySolutionMap();
 
 	@Test
@@ -54,12 +67,112 @@ public class QueryUtilsTest extends AbstractTestClass {
 		fail("bindAnon not implemented");
 	}
 
-	// ----------------------------------------------------------------------
-	// Helper methods
 	// ----------------------------------------------------------------------
 
 	private void assertBoundQueryEquals(String template, String expected) {
 		String actual = QueryUtils.bindVariables(template, bindings);
 		assertEquals("bounding results", expected, actual);
 	}
+
+	// ----------------------------------------------------------------------
+	// Test removeDuplicatesMapsFromList
+	// ----------------------------------------------------------------------
+
+	private List<Map<String, String>> theList = list(
+			map(pair("id", "1"), pair("color", "blue"), pair("size", "large")),
+			map(pair("id", "2"), pair("color", "red"), pair("size", "large"),
+					pair("parity", "odd")));
+	private List<Map<String, String>> filteredList;
+
+	@Test
+	public void noKeys() {
+		assertExpectedIDs(ids("1", "2"), keys());
+	}
+
+	@Test
+	public void emptyList() {
+		theList = new ArrayList<>();
+		assertExpectedIDs(ids(), keys("color"));
+	}
+
+	@Test
+	public void unrecognizedKey() {
+		assertExpectedIDs(ids("1", "2"), keys("bogus"));
+	}
+
+	@Test
+	public void unmatchedKey() {
+		assertExpectedIDs(ids("1", "2"), keys("parity"));
+	}
+
+	@Test
+	public void foundDuplicate() {
+		assertExpectedIDs(ids("1"), keys("size"));
+	}
+
+	@Test
+	public void noDuplicates() {
+		assertExpectedIDs(ids("1", "2"), keys("color"));
+	}
+
+	@Test
+	public void matchOneKeyOfMany() {
+		assertExpectedIDs(ids("1"), keys("color", "size"));
+	}
+
+	@Test
+	public void multipleDuplicatesOfASingleRecord() {
+		theList.add(map(pair("id", "3"), pair("size", "large")));
+		assertExpectedIDs(ids("1"), keys("color", "size"));
+	}
+
+	// ----------------------------------------------------------------------
+
+	private void assertExpectedIDs(String[] ids, String[] keys) {
+		filteredList = QueryUtils.removeDuplicatesMapsFromList(theList, keys);
+		assertEquals("ids", Arrays.asList(ids), idsInFilteredList());
+	}
+
+	private List<String> idsInFilteredList() {
+		List<String> ids = new ArrayList<>();
+		for (Map<String, String> map : filteredList) {
+			String id = map.get("id");
+			if (id == null) {
+				fail("ID was null");
+			} else {
+				ids.add(id);
+			}
+		}
+		return ids;
+	}
+
+	@SafeVarargs
+	private final List<Map<String, String>> list(Map<String, String>... maps) {
+		return new ArrayList<>(Arrays.asList(maps));
+	}
+
+	private Map<String, String> map(String[]... pairs) {
+		Map<String, String> map = new HashMap<>();
+		for (String[] pair : pairs) {
+			map.put(pair[0], pair[1]);
+		}
+		return map;
+	}
+
+	private String[] pair(String... s) {
+		return s;
+	}
+
+	private String[] keys(String... keys) {
+		return keys;
+	}
+
+	private String[] ids(String... ids) {
+		return ids;
+	}
+
+	// ----------------------------------------------------------------------
+	// Helper methods
+	// ----------------------------------------------------------------------
+
 }