RED-2845 Bugfix: Avoid the expansion if it would result in a redaction overlap and bugfix in RegExp in drools

redaction-service-v1/redaction-service-server-v1/src/main/java/com/iqser/red/service/redaction/v1/server/redaction/model/Section.java

@@ -1,13 +1,6 @@
 package com.iqser.red.service.redaction.v1.server.redaction.model;
 
-import com.iqser.red.service.redaction.v1.model.FileAttribute;
+import static com.iqser.red.service.redaction.v1.server.redaction.model.Dictionary.RECOMMENDATION_PREFIX;
-import com.iqser.red.service.redaction.v1.server.classification.model.TextBlock;
-import com.iqser.red.service.redaction.v1.server.redaction.utils.EntitySearchUtils;
-import com.iqser.red.service.redaction.v1.server.redaction.utils.Patterns;
-import lombok.Builder;
-import lombok.Data;
-import lombok.extern.slf4j.Slf4j;
-import org.apache.commons.lang3.StringUtils;
 
 import java.util.ArrayList;
 import java.util.Collection;
@@ -20,7 +13,16 @@ import java.util.regex.Matcher;
 import java.util.regex.Pattern;
 import java.util.stream.Collectors;
 
-import static com.iqser.red.service.redaction.v1.server.redaction.model.Dictionary.RECOMMENDATION_PREFIX;
+import org.apache.commons.lang3.StringUtils;
+
+import com.iqser.red.service.redaction.v1.model.FileAttribute;
+import com.iqser.red.service.redaction.v1.server.classification.model.TextBlock;
+import com.iqser.red.service.redaction.v1.server.redaction.utils.EntitySearchUtils;
+import com.iqser.red.service.redaction.v1.server.redaction.utils.Patterns;
+
+import lombok.Builder;
+import lombok.Data;
+import lombok.extern.slf4j.Slf4j;
 
 @Data
 @Slf4j
@@ -59,29 +61,57 @@ public class Section {
     private List<FileAttribute> fileAttributes = new ArrayList<>();
 
 
-    public boolean fileAttributeByIdEquals(String id, String value){
+    public boolean fileAttributeByIdEquals(String id, String value) {
-        return fileAttributes != null && fileAttributes.stream().filter(attribute -> id.equals(attribute.getId()) && value.equals(attribute.getValue())).findFirst().isPresent();
-    }
 
-    public boolean fileAttributeByPlaceholderEquals(String placeholder, String value){
+        return fileAttributes != null && fileAttributes.stream()
-        return fileAttributes != null && fileAttributes.stream().filter(attribute -> placeholder.equals(attribute.getPlaceholder()) && value.equals(attribute.getValue())).findFirst().isPresent();
+                .filter(attribute -> id.equals(attribute.getId()) && value.equals(attribute.getValue()))
-    }
+                .findFirst()
-
+                .isPresent();
-    public boolean fileAttributeByLabelEquals(String label, String value){
-        return fileAttributes != null && fileAttributes.stream().filter(attribute -> label.equals(attribute.getLabel()) && value.equals(attribute.getValue())).findFirst().isPresent();
     }
 
 
-    public boolean fileAttributeByIdEqualsIgnoreCase(String id, String value){
+    public boolean fileAttributeByPlaceholderEquals(String placeholder, String value) {
-        return fileAttributes != null && fileAttributes.stream().filter(attribute -> id.equals(attribute.getId()) && value.equalsIgnoreCase(attribute.getValue())).findFirst().isPresent();
+
+        return fileAttributes != null && fileAttributes.stream()
+                .filter(attribute -> placeholder.equals(attribute.getPlaceholder()) && value.equals(attribute.getValue()))
+                .findFirst()
+                .isPresent();
     }
 
-    public boolean fileAttributeByPlaceholderEqualsIgnoreCase(String placeholder, String value){
+
-        return fileAttributes != null && fileAttributes.stream().filter(attribute -> placeholder.equals(attribute.getPlaceholder()) && value.equalsIgnoreCase(attribute.getValue())).findFirst().isPresent();
+    public boolean fileAttributeByLabelEquals(String label, String value) {
+
+        return fileAttributes != null && fileAttributes.stream()
+                .filter(attribute -> label.equals(attribute.getLabel()) && value.equals(attribute.getValue()))
+                .findFirst()
+                .isPresent();
     }
 
-    public boolean fileAttributeByLabelEqualsIgnoreCase(String label, String value){
+
-        return fileAttributes != null && fileAttributes.stream().filter(attribute -> label.equals(attribute.getLabel()) && value.equalsIgnoreCase(attribute.getValue())).findFirst().isPresent();
+    public boolean fileAttributeByIdEqualsIgnoreCase(String id, String value) {
+
+        return fileAttributes != null && fileAttributes.stream()
+                .filter(attribute -> id.equals(attribute.getId()) && value.equalsIgnoreCase(attribute.getValue()))
+                .findFirst()
+                .isPresent();
+    }
+
+
+    public boolean fileAttributeByPlaceholderEqualsIgnoreCase(String placeholder, String value) {
+
+        return fileAttributes != null && fileAttributes.stream()
+                .filter(attribute -> placeholder.equals(attribute.getPlaceholder()) && value.equalsIgnoreCase(attribute.getValue()))
+                .findFirst()
+                .isPresent();
+    }
+
+
+    public boolean fileAttributeByLabelEqualsIgnoreCase(String label, String value) {
+
+        return fileAttributes != null && fileAttributes.stream()
+                .filter(attribute -> label.equals(attribute.getLabel()) && value.equalsIgnoreCase(attribute.getValue()))
+                .findFirst()
+                .isPresent();
     }
 
 
@@ -121,11 +151,13 @@ public class Section {
 
 
     public void expandByRegEx(String type, String pattern, boolean patternCaseInsensitive, int group) {
+
         expandByRegEx(type, pattern, patternCaseInsensitive, group, null);
     }
 
 
-    public void expandByRegEx(String type, String pattern, boolean patternCaseInsensitive, int group, String withoutPattern) {
+    public void expandByRegEx(String type, String pattern, boolean patternCaseInsensitive, int group,
+                              String withoutPattern) {
 
         Pattern compiledWithoutPattern = null;
         if (withoutPattern != null) {
@@ -141,7 +173,7 @@ public class Section {
                 continue;
             }
 
-            if(withoutPattern != null) {
+            if (withoutPattern != null) {
                 Matcher matcherWithout = compiledWithoutPattern.matcher(entity.getWord());
                 if (matcherWithout.find()) {
                     continue;
@@ -152,11 +184,10 @@ public class Section {
 
             while (matcher.find()) {
                 String match = matcher.group(group);
+
                 if (StringUtils.isNotBlank(match)) {
-
+                    Set<Entity> expandedEntities = findEntities(entity.getWord() + match, type, false, entity.isRedaction(), entity.getMatchedRule(), entity.getRedactionReason(), entity.getLegalBasis());
-                    expanded.addAll(findEntities(entity.getWord() + match, type, false, entity.isRedaction(), entity.getMatchedRule(), entity
+                    expanded.addAll(EntitySearchUtils.findNonOverlappingMatchEntities(entities, expandedEntities));
-                            .getRedactionReason(), entity.getLegalBasis()));
-
                 }
             }
         }
@@ -222,7 +253,8 @@ public class Section {
     }
 
 
-    public void ignore(String type){
+    public void ignore(String type) {
+
         entities.removeIf(entity -> entity.getType().equals(type));
     }
 

redaction-service-v1/redaction-service-server-v1/src/main/java/com/iqser/red/service/redaction/v1/server/redaction/utils/EntitySearchUtils.java

@@ -1,5 +1,16 @@
 package com.iqser.red.service.redaction.v1.server.redaction.utils;
 
+import java.util.ArrayList;
+import java.util.Comparator;
+import java.util.HashMap;
+import java.util.HashSet;
+import java.util.List;
+import java.util.Locale;
+import java.util.Map;
+import java.util.Set;
+import java.util.regex.Pattern;
+import java.util.stream.Collectors;
+
 import com.iqser.red.service.redaction.v1.server.redaction.model.Dictionary;
 import com.iqser.red.service.redaction.v1.server.redaction.model.DictionaryIncrementValue;
 import com.iqser.red.service.redaction.v1.server.redaction.model.Entity;
@@ -9,10 +20,6 @@ import com.iqser.red.service.redaction.v1.server.redaction.model.SearchableText;
 import lombok.experimental.UtilityClass;
 import lombok.extern.slf4j.Slf4j;
 
-import java.util.*;
-import java.util.regex.Pattern;
-import java.util.stream.Collectors;
-
 @Slf4j
 @UtilityClass
 @SuppressWarnings("PMD")
@@ -36,8 +43,7 @@ public class EntitySearchUtils {
                 startIndex = inputString.indexOf(cleanValue, stopIndex);
                 stopIndex = startIndex + cleanValue.length();
 
-                if (startIndex > -1 && (startIndex == 0 || Character.isWhitespace(inputString.charAt(startIndex - 1)) || isSeparator(inputString
+                if (startIndex > -1 && (startIndex == 0 || Character.isWhitespace(inputString.charAt(startIndex - 1)) || isSeparator(inputString.charAt(startIndex - 1))) && (stopIndex == inputString.length() || isSeparator(inputString.charAt(stopIndex)))) {
-                        .charAt(startIndex - 1))) && (stopIndex == inputString.length() || isSeparator(inputString.charAt(stopIndex)))) {
                     return true;
                 }
             } while (startIndex > -1);
@@ -65,8 +71,7 @@ public class EntitySearchUtils {
                 startIndex = inputString.indexOf(cleanValue, stopIndex);
                 stopIndex = startIndex + cleanValue.length();
 
-                if (startIndex > -1 && (startIndex == 0 || Character.isWhitespace(inputString.charAt(startIndex - 1)) || isSeparator(inputString
+                if (startIndex > -1 && (startIndex == 0 || Character.isWhitespace(inputString.charAt(startIndex - 1)) || isSeparator(inputString.charAt(startIndex - 1))) && (stopIndex == inputString.length() || isSeparator(inputString.charAt(stopIndex)))) {
-                        .charAt(startIndex - 1))) && (stopIndex == inputString.length() || isSeparator(inputString.charAt(stopIndex)))) {
                     found.add(new Entity(inputString.substring(startIndex, stopIndex), type, startIndex, stopIndex, headline, sectionNumber, !local, isDossierDictionary));
                 }
             } while (startIndex > -1);
@@ -120,8 +125,7 @@ public class EntitySearchUtils {
         for (Entity word : entities) {
             for (Entity inner : entities) {
                 if (inner.getWord().length() < word.getWord()
-                        .length() && inner.getStart() >= word.getStart() && inner.getEnd() <= word.getEnd() && word != inner && word
+                        .length() && inner.getStart() >= word.getStart() && inner.getEnd() <= word.getEnd() && word != inner && word.getSectionNumber() == inner.getSectionNumber()) {
-                        .getSectionNumber() == inner.getSectionNumber()) {
                     wordsToRemove.add(inner);
                 }
             }
@@ -154,4 +158,20 @@ public class EntitySearchUtils {
         entities.addAll(found);
     }
 
+
+    public Set<Entity> findNonOverlappingMatchEntities(Set<Entity> existingEntities, Set<Entity> foundEntities) {
+
+        Set<Entity> result = new HashSet<>();
+        if (existingEntities != null && foundEntities != null) {
+            for (Entity existingEntity : existingEntities) {
+                for (Entity foundEntity : foundEntities) {
+                    if (existingEntity.getEnd() < foundEntity.getStart() || foundEntity.getEnd() < existingEntity.getStart()) {
+                        result.add(foundEntity);
+                    }
+                }
+            }
+        }
+        return result;
+    }
+
 }

redaction-service-v1/redaction-service-server-v1/src/test/java/com/iqser/red/service/redaction/v1/server/redaction/utils/EntitySearchUtilsTest.java

@@ -0,0 +1,173 @@
+package com.iqser.red.service.redaction.v1.server.redaction.utils;
+
+import static org.assertj.core.api.Assertions.assertThat;
+
+import java.util.HashSet;
+import java.util.Set;
+
+import org.junit.Test;
+
+import com.iqser.red.service.redaction.v1.server.redaction.model.Entity;
+
+public class EntitySearchUtilsTest {
+
+    /*
+     * Text:     Batman X. Superman Y.
+     * Position: 0123456789
+     *                     0123456789
+     *                               0123456789
+     */
+    @Test
+    public void testNotOverlappingEntitiesExpandedEnd() {
+
+        // Arrange
+        Set<Entity> existingEntities = new HashSet<>();
+        Entity existingEntity1 = new Entity("Batman", "fake type", 0, 5, "fake headline", 0, false, false);
+        Entity existingEntity2 = new Entity("Superman", "fake type", 10, 17, "fake headline", 0, false, false);
+        existingEntities.add(existingEntity1);
+        existingEntities.add(existingEntity2);
+
+        Set<Entity> foundEntities = new HashSet<>();
+        Entity foundEntities1 = new Entity("Batman X.", "fake type", 0, 8, "fake headline", 0, false, false);
+        Entity foundEntities2 = new Entity("Superman Y.", "fake type", 10, 20, "fake headline", 0, false, false);
+        foundEntities.add(foundEntities1);
+        foundEntities.add(foundEntities2);
+
+        // Act
+        Set<Entity> result = EntitySearchUtils.findNonOverlappingMatchEntities(existingEntities, foundEntities);
+
+        // Assert
+        assertThat(result.size()).isEqualTo(2);
+        assertThat(result).contains(foundEntities1);
+        assertThat(result).contains(foundEntities2);
+
+    }
+
+
+    /*
+     * Text:     Batman X. Superman Y.
+     * Position: 0123456789
+     *                     0123456789
+     *                               0123456789
+     */
+    @Test
+    public void testNotOverlappingEntitiesExpandedStartAndEndOverlapping() {
+
+        // Arrange
+        Set<Entity> existingEntities = new HashSet<>();
+        Entity existingEntity1 = new Entity("Batman X.", "fake type", 0, 8, "fake headline", 0, false, false);
+        Entity existingEntity2 = new Entity("Superman", "fake type", 10, 17, "fake headline", 0, false, false);
+        existingEntities.add(existingEntity1);
+        existingEntities.add(existingEntity2);
+
+        Set<Entity> foundEntities = new HashSet<>();
+        Entity foundEntities1 = new Entity("Batman X.", "fake type", 0, 8, "fake headline", 0, false, false);
+        Entity foundEntities2 = new Entity("X. Superman Y.", "fake type", 7, 20, "fake headline", 0, false, false);
+        foundEntities.add(foundEntities1);
+        foundEntities.add(foundEntities2);
+
+        // Act
+        Set<Entity> result = EntitySearchUtils.findNonOverlappingMatchEntities(existingEntities, foundEntities);
+
+        // Assert
+        assertThat(result.size()).isEqualTo(1);
+        assertThat(result).contains(foundEntities1);
+
+    }
+
+
+    /*
+     * Text:     Batman X. Superman Y.
+     * Position: 0123456789
+     *                     0123456789
+     *                               0123456789
+     */
+    @Test
+    public void testNotOverlappingEntitiesExpandedStartAndEnd() {
+
+        // Arrange
+        Set<Entity> existingEntities = new HashSet<>();
+        Entity existingEntity1 = new Entity("Batman X.", "fake type", 0, 8, "fake headline", 0, false, false);
+        Entity existingEntity2 = new Entity("Superman", "fake type", 10, 17, "fake headline", 0, false, false);
+        existingEntities.add(existingEntity1);
+        existingEntities.add(existingEntity2);
+
+        Set<Entity> foundEntities = new HashSet<>();
+        Entity foundEntities1 = new Entity("Batman X.", "fake type", 0, 8, "fake headline", 0, false, false);
+        Entity foundEntities2 = new Entity("X. Superman", "fake type", 7, 17, "fake headline", 0, false, false);
+        foundEntities.add(foundEntities1);
+        foundEntities.add(foundEntities2);
+
+        // Act
+        Set<Entity> result = EntitySearchUtils.findNonOverlappingMatchEntities(existingEntities, foundEntities);
+
+        // Assert
+        assertThat(result.size()).isEqualTo(1);
+        assertThat(result).contains(foundEntities1);
+
+    }
+
+
+    /*
+     * Text:     Batman X. Superman Y.
+     * Position: 0123456789
+     *                     0123456789
+     *                               0123456789
+     */
+    @Test
+    public void testNotOverlappingEntitiesExpandedExistingAndExpandedEnd() {
+
+        // Arrange
+        Set<Entity> existingEntities = new HashSet<>();
+        Entity existingEntity1 = new Entity("X. Superman", "fake type", 7, 17, "fake headline", 0, false, false);
+        Entity existingEntity2 = new Entity("Batman", "fake type", 0, 5, "fake headline", 0, false, false);
+        existingEntities.add(existingEntity1);
+        existingEntities.add(existingEntity2);
+
+        Set<Entity> foundEntities = new HashSet<>();
+        Entity foundEntities1 = new Entity("Batman X.", "fake type", 0, 8, "fake headline", 0, false, false);
+        Entity foundEntities2 = new Entity("Superman", "fake type", 10, 17, "fake headline", 0, false, false);
+        foundEntities.add(foundEntities1);
+        foundEntities.add(foundEntities2);
+
+        // Act
+        Set<Entity> result = EntitySearchUtils.findNonOverlappingMatchEntities(existingEntities, foundEntities);
+
+        // Assert
+        assertThat(result.size()).isEqualTo(1);
+        assertThat(result).contains(foundEntities2);
+
+    }
+
+    /*
+     * Text:     Batman X. Superman Y.
+     * Position: 0123456789
+     *                     0123456789
+     *                               0123456789
+     */
+    @Test
+    public void testNotOverlappingEntitiesExpandedEndLong() {
+
+        // Arrange
+        Set<Entity> existingEntities = new HashSet<>();
+        Entity existingEntity1 = new Entity("X. Superman", "fake type", 7, 17, "fake headline", 0, false, false);
+        Entity existingEntity2 = new Entity("Batman", "fake type", 0, 5, "fake headline", 0, false, false);
+        existingEntities.add(existingEntity1);
+        existingEntities.add(existingEntity2);
+
+        Set<Entity> foundEntities = new HashSet<>();
+        Entity foundEntities1 = new Entity("Batman X. Superman", "fake type", 0, 17, "fake headline", 0, false, false);
+        Entity foundEntities2 = new Entity("Superman", "fake type", 10, 17, "fake headline", 0, false, false);
+        foundEntities.add(foundEntities1);
+        foundEntities.add(foundEntities2);
+
+        // Act
+        Set<Entity> result = EntitySearchUtils.findNonOverlappingMatchEntities(existingEntities, foundEntities);
+
+        // Assert
+        assertThat(result.size()).isEqualTo(1);
+        assertThat(result).contains(foundEntities2);
+
+    }
+
+}

redaction-service-v1/redaction-service-server-v1/src/test/resources/drools/rules.drl

@@ -19,8 +19,8 @@ rule "0: Expand CBI Authors with firstname initials"
     when
         Section(matchesType("CBI_author") || matchesType("recommendation_CBI_author"))
     then
-        section.expandByRegEx("CBI_author", "(,? [A-Z]\\.?( ?[A-Z]\\.?)?( ?[A-Z]\\.?)?\\b\\.?)", false, 1, "[^\\s]+");
+        section.expandByRegEx("CBI_author", "(,? [A-Z]\\.?( ?[A-Z]\\.?)?( ?[A-Z]\\.?)?\\b\\.?)", false, 1, "[\\s]+");
-        section.expandByRegEx("recommendation_CBI_author", "(,? [A-Z]\\.?( ?[A-Z]\\.?)?( ?[A-Z]\\.?)?\\b\\.?)", false, 1, "[^\\s]+");
+        section.expandByRegEx("recommendation_CBI_author", "(,? [A-Z]\\.?( ?[A-Z]\\.?)?( ?[A-Z]\\.?)?\\b\\.?)", false, 1, "[\\s]+");
     end