Merge branch 'feature/coref-resolution'

connection-generator/src/main/java/edu/kit/kastel/mcse/ardoco/core/connectiongenerator/extractors/InitialConnectionAgent.java → connection-generator/src/main/java/edu/kit/kastel/mcse/ardoco/core/connectiongenerator/agents/InitialConnectionAgent.java

@@ -1,4 +1,4 @@
-package edu.kit.kastel.mcse.ardoco.core.connectiongenerator.extractors;
+package edu.kit.kastel.mcse.ardoco.core.connectiongenerator.agents;
 
 import java.util.Map;
 

connection-generator/src/main/java/edu/kit/kastel/mcse/ardoco/core/connectiongenerator/extractors/ExtractedTermsExtractor.java

@@ -105,7 +105,7 @@ private void createRecommendedInstancesForTerm(IWord node) {
 
     }
 
-    private ImmutableList<ITermMapping> getPossibleOccurredTermMappingsToThisSpot(ImmutableList<ITermMapping> termMappings, IWord n) {
+    private static ImmutableList<ITermMapping> getPossibleOccurredTermMappingsToThisSpot(ImmutableList<ITermMapping> termMappings, IWord n) {
 
         MutableList<ITermMapping> possibleOccuredTermMappings = Lists.mutable.empty();
         String word = n.getText();
@@ -133,11 +133,15 @@ private ImmutableList<ITermMapping> getPossibleOccurredTermMappingsToThisSpot(Im
 
     }
 
-    private boolean areNextWordsEqualToReferences(ImmutableList<INounMapping> references, IWord currentWord, int currentPosition) {
+    private static boolean areNextWordsEqualToReferences(ImmutableList<INounMapping> references, IWord currentWord, int currentPosition) {
         var stop = false;
 
         for (int i = currentPosition + 1; i < references.size() && !stop; i++) {
-            String postWord = currentWord.getNextWord().getText();
+            var nextWord = currentWord.getNextWord();
+            if (nextWord == null) {
+                continue;
+            }
+            String postWord = nextWord.getText();
             String reference = references.get(i).getReference();
 
             if (!SimilarityUtils.areWordsSimilar(reference, postWord)) {
@@ -148,11 +152,15 @@ private boolean areNextWordsEqualToReferences(ImmutableList<INounMapping> refere
         return stop;
     }
 
-    private boolean arePreviousWordsEqualToReferences(ImmutableList<INounMapping> references, IWord currentWord, int currentPosition) {
+    private static boolean arePreviousWordsEqualToReferences(ImmutableList<INounMapping> references, IWord currentWord, int currentPosition) {
         var stop = false;
 
         for (int i = currentPosition - 1; i >= 0 && !stop; i--) {
-            String preWord = currentWord.getPreWord().getText();
+            IWord previousWord = currentWord.getPreWord();
+            if (previousWord == null) {
+                continue;
+            }
+            String preWord = previousWord.getText();
             String reference = references.get(i).getReference();
             if (!SimilarityUtils.areWordsSimilar(reference, preWord)) {
                 stop = true;
@@ -170,7 +178,7 @@ private void createRecommendedInstancesForAdjacentTermMappings(IWord termStartNo
         createRecommendedInstancesOfAdjacentTerms(term, adjCompleteTermMappings.toImmutable());
     }
 
-    private ImmutableList<ITermMapping> getCompletePreAdjTermMappings(ImmutableList<ITermMapping> possibleTermMappings, IWord termStartNode) {
+    private static ImmutableList<ITermMapping> getCompletePreAdjTermMappings(ImmutableList<ITermMapping> possibleTermMappings, IWord termStartNode) {
         int sentence = termStartNode.getSentenceNo();
         IWord preTermNode = termStartNode.getPreWord();
 
@@ -183,7 +191,7 @@ private ImmutableList<ITermMapping> getCompletePreAdjTermMappings(ImmutableList<
             while (!nounMappings.isEmpty()) {
                 INounMapping resultOfPreMatch = matchNode(nounMappings.toImmutable(), preTermNode);
 
-                if (sentence == preTermNode.getSentenceNo() || resultOfPreMatch == null) {
+                if (preTermNode == null || sentence == preTermNode.getSentenceNo() || resultOfPreMatch == null) {
                     break;
                 }
 
@@ -199,7 +207,7 @@ private ImmutableList<ITermMapping> getCompletePreAdjTermMappings(ImmutableList<
         return adjCompleteTermMappings.toImmutable();
     }
 
-    private ImmutableList<ITermMapping> getCompleteAfterAdjTermMappings(ImmutableList<ITermMapping> possibleTermMappings, IWord termStartNode,
+    private static ImmutableList<ITermMapping> getCompleteAfterAdjTermMappings(ImmutableList<ITermMapping> possibleTermMappings, IWord termStartNode,
             ITermMapping term) {
 
         int sentence = termStartNode.getSentenceNo();
@@ -235,7 +243,10 @@ private ImmutableList<ITermMapping> getCompleteAfterAdjTermMappings(ImmutableLis
 
     }
 
-    private INounMapping matchNode(ImmutableList<INounMapping> nounMappings, IWord node) {
+    private static INounMapping matchNode(ImmutableList<INounMapping> nounMappings, IWord node) {
+        if (node == null) {
+            return null;
+        }
         for (INounMapping mapping : nounMappings) {
             if (mapping.getWords().contains(node)) {
                 return mapping;
@@ -244,7 +255,7 @@ private INounMapping matchNode(ImmutableList<INounMapping> nounMappings, IWord n
         return null;
     }
 
-    private IWord getAfterTermNode(IWord termStartNode, ITermMapping term) {
+    private static IWord getAfterTermNode(IWord termStartNode, ITermMapping term) {
         IWord afterTermNode = termStartNode.getNextWord();
         for (var i = 0; i < term.getMappings().size() && afterTermNode != null; i++) {
             afterTermNode = afterTermNode.getNextWord();
@@ -264,11 +275,11 @@ private ImmutableList<INounMapping> getTermAdjacentNounMappings(IWord node, ITer
 
         if (preTermNode != null && sentence == preTermNode.getSentenceNo()) {
 
-            ImmutableList<INounMapping> nounMappingsOfPreTermNode = textState.getNounMappingsByNode(preTermNode);
+            ImmutableList<INounMapping> nounMappingsOfPreTermNode = textState.getNounMappingsByWord(preTermNode);
             possibleMappings.addAll(nounMappingsOfPreTermNode.castToCollection());
         }
         if (afterTermNode != null && sentence == afterTermNode.getSentenceNo()) {
-            ImmutableList<INounMapping> nounMappingsOfAfterTermNode = textState.getNounMappingsByNode(afterTermNode);
+            ImmutableList<INounMapping> nounMappingsOfAfterTermNode = textState.getNounMappingsByWord(afterTermNode);
             possibleMappings.addAll(nounMappingsOfAfterTermNode.castToCollection());
         }
         return possibleMappings.select(nounMapping -> nounMapping.getKind() != kind).toImmutable();

connection-generator/src/main/java/edu/kit/kastel/mcse/ardoco/core/datastructures/InstanceLink.java

@@ -110,7 +110,7 @@ public String getNameOccurencesAsString() {
         Set<String> names = new HashSet<>();
         MutableList<Integer> namePositions = Lists.mutable.empty();
         for (INounMapping nameMapping : textualInstance.getNameMappings()) {
-            names.addAll(nameMapping.getOccurrences().castToCollection());
+            names.addAll(nameMapping.getSurfaceForms().castToCollection());
             namePositions.addAll(nameMapping.getMappingSentenceNo().castToCollection());
         }
 
@@ -126,11 +126,11 @@ public String toString() {
         MutableList<Integer> typePositions = Lists.mutable.empty();
 
         for (INounMapping nameMapping : textualInstance.getNameMappings()) {
-            names.addAll(nameMapping.getOccurrences().castToCollection());
+            names.addAll(nameMapping.getSurfaceForms().castToCollection());
             namePositions.addAll(nameMapping.getMappingSentenceNo().castToCollection());
         }
         for (INounMapping typeMapping : textualInstance.getTypeMappings()) {
-            types.addAll(typeMapping.getOccurrences().castToCollection());
+            types.addAll(typeMapping.getSurfaceForms().castToCollection());
             typePositions.addAll(typeMapping.getMappingSentenceNo().castToCollection());
         }
         return "InstanceMapping [ uid=" + modelInstance.getUid() + ", name=" + modelInstance.getLongestName() + //

data-structures/src/main/java/edu/kit/kastel/mcse/ardoco/core/datastructures/agents/Configuration.java

@@ -2,6 +2,7 @@
 
 import java.io.File;
 import java.io.IOException;
+import java.nio.charset.StandardCharsets;
 import java.util.Map;
 import java.util.Scanner;
 
@@ -54,6 +55,19 @@ protected static double getPropertyAsDouble(String key, Map<String, String> conf
         return Double.parseDouble(configs.get(key));
     }
 
+    /**
+     * Returns the specified property of the config file as a boolean if it is set.
+     *
+     * @param key     name of the specified property
+     * @param configs the configuration map
+     * @return value of the property as a boolean. True, if the value for the key is "true", "yes", or "1" ignoring
+     *         case.
+     */
+    protected static boolean isPropertyEnabled(String key, Map<String, String> configs) {
+        var propValue = configs.get(key).strip();
+        return Boolean.parseBoolean(propValue) || propValue.equalsIgnoreCase("yes") || propValue.equalsIgnoreCase("1");
+    }
+
     /**
      * Get all properties and raw values of a configuration
      *
@@ -84,21 +98,19 @@ public static void mergeConfigToMap(Map<String, String> configs, Configuration c
      * @param additionalConfigs the file with additional configs
      */
     public static void overrideConfigInMap(Map<String, String> configs, File additionalConfigs) {
-        try (var scan = new Scanner(additionalConfigs)) {
+        try (var scan = new Scanner(additionalConfigs, StandardCharsets.UTF_8)) {
             while (scan.hasNextLine()) {
                 String line = scan.nextLine();
+                String[] kv = new String[0];
                 if (line == null || line.isBlank()) {
                     logger.warn("Illegal Line in config: \"{}\"", line);
-                    continue;
+                } else {
+                    kv = line.trim().split("=", 2);
                 }
 
-                String[] kv = line.trim().split("=", 2);
                 if (kv.length != 2) {
                     logger.warn("Illegal Line in config: \"{}\"", line);
-                    continue;
-                }
-
-                if (configs.containsKey(kv[0].trim())) {
+                } else if (configs.containsKey(kv[0].trim())) {
                     configs.put(kv[0].trim(), kv[1].trim());
                 }
             }

data-structures/src/main/java/edu/kit/kastel/mcse/ardoco/core/datastructures/definitions/IConnectionState.java

@@ -1,8 +1,9 @@
 package edu.kit.kastel.mcse.ardoco.core.datastructures.definitions;
 
-import org.eclipse.collections.api.factory.Lists;
+import org.eclipse.collections.api.factory.Sets;
 import org.eclipse.collections.api.list.ImmutableList;
-import org.eclipse.collections.api.list.MutableList;
+import org.eclipse.collections.api.set.ImmutableSet;
+import org.eclipse.collections.api.set.MutableSet;
 
 import edu.kit.kastel.mcse.ardoco.core.datastructures.modules.IState;
 
@@ -56,14 +57,19 @@ public interface IConnectionState extends IState<IConnectionState> {
      *
      * @return list of tracelinks within this connection state
      */
-    default ImmutableList<Tracelink> getTraceLinks() {
-        MutableList<Tracelink> tracelinks = Lists.mutable.empty();
+    default ImmutableSet<Tracelink> getTraceLinks() {
+        MutableSet<Tracelink> tracelinks = Sets.mutable.empty();
         for (var instanceLink : getInstanceLinks()) {
-            for (var nm : instanceLink.getTextualInstance().getNameMappings()) {
+            var textualInstance = instanceLink.getTextualInstance();
+            for (var nm : textualInstance.getNameMappings()) {
                 for (var word : nm.getWords()) {
                     var tracelink = new Tracelink(instanceLink, instanceLink.getModelInstance(), word);
                     tracelinks.add(tracelink);
                 }
+                for (var word : nm.getCoreferences()) {
+                    var tracelink = new Tracelink(instanceLink, instanceLink.getModelInstance(), word);
+                    tracelinks.add(tracelink);
+                }
             }
         }
         return tracelinks.toImmutable();

data-structures/src/main/java/edu/kit/kastel/mcse/ardoco/core/datastructures/definitions/ICorefCluster.java

@@ -0,0 +1,32 @@
+/**
+ *
+ */
+package edu.kit.kastel.mcse.ardoco.core.datastructures.definitions;
+
+import java.util.StringJoiner;
+
+import org.eclipse.collections.api.list.ImmutableList;
+
+/**
+ * This defines the interface of a Coreference Cluster (CorefCluster). A CorefCluster is a cluster that collects all
+ * mentions of an entity and has a representative mention.
+ *
+ * @author Jan Keim
+ *
+ */
+public interface ICorefCluster {
+
+    int getId();
+
+    String getRepresentativeMention();
+
+    ImmutableList<ImmutableList<IWord>> getMentions();
+
+    static String getTextForMention(ImmutableList<IWord> mention) {
+        var textJoiner = new StringJoiner(" ");
+        for (var word : mention) {
+            textJoiner.add(word.getText());
+        }
+        return textJoiner.toString();
+    }
+}

data-structures/src/main/java/edu/kit/kastel/mcse/ardoco/core/datastructures/definitions/INounMapping.java

@@ -1,5 +1,6 @@
 package edu.kit.kastel.mcse.ardoco.core.datastructures.definitions;
 
+import java.util.Collection;
 import java.util.Map;
 
 import org.eclipse.collections.api.list.ImmutableList;
@@ -27,16 +28,16 @@ public interface INounMapping extends ICopyable<INounMapping> {
     void hardSetProbability(double probability);
 
     /**
-     * Returns the occurrences of this mapping.
+     * Returns the surface forms (previously called occurrences) of this mapping.
      *
      * @return all appearances of the mapping
      */
-    ImmutableList<String> getOccurrences();
+    ImmutableList<String> getSurfaceForms();
 
     /**
-     * Returns all nodes contained by the mapping.
+     * Returns all words that are contained by the mapping. This should include coreferences.
      *
-     * @return all mapping nodes
+     * @return all words that are referenced with this mapping
      */
     ImmutableList<IWord> getWords();
 
@@ -45,14 +46,14 @@ public interface INounMapping extends ICopyable<INounMapping> {
      *
      * @param nodes graph nodes to add to the mapping
      */
-    void addNodes(ImmutableList<IWord> nodes);
+    void addWords(ImmutableList<IWord> nodes);
 
     /**
      * Adds a node to the mapping, it its not already contained.
      *
      * @param n graph node to add.
      */
-    void addNode(IWord n);
+    void addWord(IWord n);
 
     /**
      * Returns the probability of being a mapping of its kind.
@@ -148,4 +149,36 @@ public interface INounMapping extends ICopyable<INounMapping> {
      * @return the distribution
      */
     Map<MappingKind, Double> getDistribution();
+
+    /**
+     * @param coreference the coreference to add
+     */
+    void addCoreference(IWord coreference);
+
+    /**
+     * @param coreferences the coreferences to add
+     */
+    void addCoreferences(Collection<IWord> coreferences);
+
+    /**
+     * @return the coreferences
+     */
+    ImmutableList<IWord> getCoreferences();
+
+    /**
+     * Creates a new INounMapping that resutls when merging the data from the INounMapping with a given other
+     * INounMapping
+     *
+     * @param other the other INounMapping
+     * @return new INounMapping that is a merge of the given INounMappings
+     */
+    INounMapping merge(INounMapping other);
+
+    /**
+     * Adds the kind with probability.
+     *
+     * @param kind        the kind
+     * @param probability the probability
+     */
+    void addKindWithProbability(MappingKind kind, double probability);
 }

data-structures/src/main/java/edu/kit/kastel/mcse/ardoco/core/datastructures/definitions/IText.java

@@ -29,4 +29,11 @@ default int getLength() {
      * @return the words
      */
     ImmutableList<IWord> getWords();
+
+    /**
+     * Get all Coreference Clusters that the text has
+     *
+     * @return the coreference clusters
+     */
+    ImmutableList<ICorefCluster> getCorefClusters();
 }