RED-6009 - Document Tree Structure

* added gaps in Redaction Positions, if they are farther than five character widths apart
2023-05-10 16:45:23 +02:00 · 2023-05-10 10:41:59 +02:00 · 2023-05-09 17:20:33 +02:00 · 2023-05-09 16:50:52 +02:00 · 2023-05-09 14:32:53 +02:00 · 2023-05-09 14:12:12 +02:00
204 changed files with 41706 additions and 7528 deletions
@@ -144,7 +144,7 @@ public class PlanSpec {
                                        .inlineBody("#!/bin/bash\n" + "set -e\n" + "rm -rf ./*"),
                                new VcsCheckoutTask().description("Checkout Default Repository").cleanCheckout(true).checkoutItems(new CheckoutItem().defaultRepository()),
                                new ScriptTask().description("Sonar").location(Location.FILE).fileFromPath("bamboo-specs/src/main/resources/scripts/sonar-java.sh").argument(SERVICE_NAME))
-                        .dockerConfiguration(new DockerConfiguration().image("nexus.iqser.com:5001/infra/maven:3.8.4-openjdk-17-slim")
+                        .dockerConfiguration(new DockerConfiguration().image("nexus.knecon.com:5001/infra/maven:3.8.4-openjdk-17-slim")
                                .dockerRunArguments("--net=host")
                                .volume("/etc/maven/settings.xml", "/usr/share/maven/conf/settings.xml")
                                .volume("/var/run/docker.sock", "/var/run/docker.sock"))))
@@ -23,6 +23,7 @@

    <properties>
        <pdfbox.version>2.0.24</pdfbox.version>
+        <lombok.version>1.18.26</lombok.version>
    </properties>


@@ -88,5 +89,26 @@
                </plugin>
            </plugins>
        </pluginManagement>
+        <plugins>
+            <plugin>
+                <groupId>org.projectlombok</groupId>
+                <artifactId>lombok-maven-plugin</artifactId>
+                <version>1.18.20.0</version>
+                <executions>
+                    <execution>
+                        <id>delombok</id>
+                        <phase>generate-sources</phase>
+                        <goals>
+                            <goal>delombok</goal>
+                        </goals>
+                        <configuration>
+                            <addOutputDirectory>false</addOutputDirectory>
+                            <sourceDirectory>src/main/java</sourceDirectory>
+                            <outputDirectory>${delomboked.sources}</outputDirectory>
+                        </configuration>
+                    </execution>
+                </executions>
+            </plugin>
+        </plugins>
    </build>
 </project>
@@ -39,7 +39,6 @@
                </exclusion>
            </exclusions>
        </dependency>
-
    </dependencies>

    <build>
@@ -12,7 +12,7 @@
    <artifactId>redaction-service-server-v1</artifactId>

    <properties>
-        <drools.version>7.73.0.Final</drools.version>
+        <drools.version>8.37.0.Final</drools.version>
        <kie.version>7.73.0.Final</kie.version>
        <locationtech.version>1.19.0</locationtech.version>
        <javaassist.version>3.29.2-GA</javaassist.version>
@@ -64,7 +64,12 @@
        </dependency>
        <dependency>
            <groupId>org.drools</groupId>
-            <artifactId>drools-core</artifactId>
+            <artifactId>drools-engine</artifactId>
+            <version>${drools.version}</version>
+        </dependency>
+        <dependency>
+            <groupId>org.drools</groupId>
+            <artifactId>drools-mvel</artifactId>
            <version>${drools.version}</version>
        </dependency>
        <dependency>
@@ -198,5 +203,4 @@
            </plugin>
        </plugins>
    </build>
-
 </project>
@@ -1,8 +0,0 @@
-package com.iqser.red.service.redaction.v1.server.classification.model;
-
-public enum Orientation {
-
-    NONE,
-    LEFT,
-    RIGHT
-}
@@ -1,65 +0,0 @@
-package com.iqser.red.service.redaction.v1.server.classification.model;
-
-import com.iqser.red.service.redaction.v1.server.redaction.model.PdfImage;
-import com.iqser.red.service.redaction.v1.server.redaction.model.SearchableText;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.AbstractTextContainer;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.Table;
-
-import lombok.Data;
-import lombok.NoArgsConstructor;
-
-import java.util.ArrayList;
-import java.util.List;
-
-@Data
-@NoArgsConstructor
-public class Paragraph implements Comparable {
-
-    private List<AbstractTextContainer> pageBlocks = new ArrayList<>();
-    private List<PdfImage> images = new ArrayList<>();
-    private String headline;
-
-
-    public SearchableText getSearchableText() {
-
-        SearchableText searchableText = new SearchableText();
-        pageBlocks.forEach(block -> {
-            if (block instanceof TextBlock) {
-                searchableText.addAll(((TextBlock) block).getSequences());
-            }
-        });
-        return searchableText;
-    }
-
-
-    public List<Table> getTables() {
-
-        List<Table> tables = new ArrayList<>();
-        pageBlocks.forEach(block -> {
-            if (block instanceof Table) {
-                tables.add((Table) block);
-            }
-        });
-        return tables;
-    }
-
-
-    public List<TextBlock> getTextBlocks() {
-
-        List<TextBlock> textBlocks = new ArrayList<>();
-        pageBlocks.forEach(block -> {
-            if (block instanceof TextBlock) {
-                textBlocks.add((TextBlock) block);
-            }
-        });
-        return textBlocks;
-    }
-
-
-    @Override
-    public int compareTo(Object o) {
-
-        return 0;
-    }
-
-}
@@ -1,64 +0,0 @@
-package com.iqser.red.service.redaction.v1.server.classification.model;
-
-import java.util.ArrayList;
-import java.util.HashMap;
-import java.util.HashSet;
-import java.util.List;
-import java.util.Map;
-import java.util.Set;
-
-import com.fasterxml.jackson.annotation.JsonIgnore;
-import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.section.SectionArea;
-import com.iqser.red.service.redaction.v1.server.redaction.model.CellValue;
-import com.iqser.red.service.redaction.v1.server.redaction.model.Image;
-import com.iqser.red.service.redaction.v1.server.redaction.model.SearchableText;
-
-import lombok.AllArgsConstructor;
-import lombok.Builder;
-import lombok.Data;
-import lombok.NoArgsConstructor;
-
-@Data
-@Builder
-@NoArgsConstructor
-@AllArgsConstructor
-public class SectionText {
-
-    private int sectionNumber;
-    private String text;
-
-    private boolean isTable;
-    private String headline;
-
-    @Builder.Default
-    private List<SectionArea> sectionAreas = new ArrayList<>();
-    @Builder.Default
-    private Set<Image> images = new HashSet<>();
-    @Builder.Default
-    private List<TextBlock> textBlocks = new ArrayList<>();
-    @Builder.Default
-    private Map<String, CellValue> tabularData = new HashMap<>();
-    @Builder.Default
-    private List<Integer> cellStarts = new ArrayList<>();
-
-
-    public void setTabularData(Map<String, CellValue> tabularData) {
-
-        tabularData.remove(null);
-        this.tabularData = tabularData;
-    }
-
-
-    @JsonIgnore
-    public SearchableText getSearchableText() {
-
-        SearchableText searchableText = new SearchableText();
-        textBlocks.forEach(block -> {
-            if (block != null) {
-                searchableText.addAll(block.getSequences());
-            }
-        });
-        return searchableText;
-    }
-
-}
@@ -1,19 +0,0 @@
-package com.iqser.red.service.redaction.v1.server.classification.model;
-
-
-import lombok.AllArgsConstructor;
-import lombok.Data;
-import lombok.NoArgsConstructor;
-
-import java.util.ArrayList;
-import java.util.List;
-
-@Data
-@NoArgsConstructor
-@AllArgsConstructor
-public class Text {
-
-    private int numberOfPages;
-    private List<SectionText> sectionTexts = new ArrayList<>();
-
-}
@@ -10,7 +10,7 @@ import lombok.NoArgsConstructor;
@Builder
@AllArgsConstructor
@NoArgsConstructor
-public class EntityRecogintionEntity {
+public class EntityRecognitionEntity {

    private String value;
    private int startOffset;
@@ -4,7 +4,6 @@ import java.util.HashMap;
 import java.util.List;
 import java.util.Map;

-
 import lombok.AllArgsConstructor;
 import lombok.Data;
 import lombok.NoArgsConstructor;
@@ -12,8 +11,8 @@ import lombok.NoArgsConstructor;
@Data
@NoArgsConstructor
@AllArgsConstructor
-public class NerEntities {
+public class NerEntitiesModel {

-    private Map<Integer, List<EntityRecogintionEntity>> data = new HashMap<>();
+    private Map<Integer, List<EntityRecognitionEntity>> data = new HashMap<>();

 }
@@ -1,13 +1,13 @@
 package com.iqser.red.service.redaction.v1.server.controller;

+import org.springframework.web.bind.annotation.RestController;
+
 import com.iqser.red.service.redaction.v1.model.RuleBuilderModel;
 import com.iqser.red.service.redaction.v1.resources.RuleBuilderResource;
-import com.iqser.red.service.redaction.v1.server.redaction.rulebuilder.RuleBuilderModelService;
+import com.iqser.red.service.redaction.v1.server.redaction.service.RuleBuilderModelService;

 import lombok.RequiredArgsConstructor;

-import org.springframework.web.bind.annotation.RestController;
-
@RestController
@RequiredArgsConstructor
 public class RuleBuilderController implements RuleBuilderResource {
@@ -1,5 +1,6 @@
-package com.iqser.red.service.redaction.v1.server.segmentation;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter;

+import java.awt.geom.Rectangle2D;
 import java.util.ArrayList;
 import java.util.HashMap;
 import java.util.List;
@@ -10,11 +11,10 @@ import org.springframework.stereotype.Service;

 import com.fasterxml.jackson.databind.ObjectMapper;
 import com.iqser.red.service.persistence.service.v1.api.shared.model.dossiertemplate.dossier.file.FileType;
-import com.iqser.red.service.redaction.v1.server.classification.model.Page;
-import com.iqser.red.service.redaction.v1.server.redaction.model.ImageType;
-import com.iqser.red.service.redaction.v1.server.redaction.model.PdfImage;
-import com.iqser.red.service.redaction.v1.server.redaction.model.RedRectangle2D;
-import com.iqser.red.service.redaction.v1.server.redaction.model.image.ImageServiceResponse;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.image.ImageServiceResponse;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.ClassificationPage;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.image.ClassifiedImage;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.ImageType;
 import com.iqser.red.service.redaction.v1.server.storage.RedactionStorageService;

 import lombok.RequiredArgsConstructor;
@@ -22,26 +22,26 @@ import lombok.SneakyThrows;

@Service
@RequiredArgsConstructor
-public class ImageService {
+public class ImageServiceResponseAdapter {

    private final ObjectMapper objectMapper;
    private final RedactionStorageService redactionStorageService;


    @SneakyThrows
-    public Map<Integer, List<PdfImage>> convertImages(String dossierId, String fileId) {
+    public Map<Integer, List<ClassifiedImage>> convertImages(String dossierId, String fileId) {

        var imageClassificationStream = redactionStorageService.getStoredObject(RedactionStorageService.StorageIdUtils.getStorageId(dossierId, fileId, FileType.IMAGE_INFO));

        ImageServiceResponse imageServiceResponse = objectMapper.readValue(imageClassificationStream, ImageServiceResponse.class);

-        Map<Integer, List<PdfImage>> images = new HashMap<>();
+        Map<Integer, List<ClassifiedImage>> images = new HashMap<>();
        imageServiceResponse.getData().forEach(imageMetadata -> {
            var classification = imageMetadata.getFilters().isAllPassed() ? ImageType.valueOf(imageMetadata.getClassification()
                    .getLabel()
                    .toUpperCase(Locale.ROOT)) : ImageType.OTHER;
            images.computeIfAbsent(imageMetadata.getPosition().getPageNumber(), x -> new ArrayList<>())
-                    .add(new PdfImage(new RedRectangle2D(imageMetadata.getPosition().getX1(),
+                    .add(new ClassifiedImage(new Rectangle2D.Double(imageMetadata.getPosition().getX1(),
                            imageMetadata.getPosition().getY1(),
                            imageMetadata.getGeometry().getWidth(),
                            imageMetadata.getGeometry().getHeight()), classification, imageMetadata.isAlpha(), imageMetadata.getPosition().getPageNumber()));
@@ -53,7 +53,7 @@ public class ImageService {
                    .getLabel()
                    .toUpperCase(Locale.ROOT)) : ImageType.OTHER;
            images.computeIfAbsent(imageMetadata.getPosition().getPageNumber(), x -> new ArrayList<>())
-                    .add(new PdfImage(new RedRectangle2D(imageMetadata.getPosition().getX1(),
+                    .add(new ClassifiedImage(new Rectangle2D.Double(imageMetadata.getPosition().getX1(),
                            imageMetadata.getPosition().getY1(),
                            imageMetadata.getGeometry().getWidth(),
                            imageMetadata.getGeometry().getHeight()), classification, imageMetadata.isAlpha(), imageMetadata.getPosition().getPageNumber()));
@@ -63,7 +63,7 @@ public class ImageService {
    }


-    public void findOcr(Page page) {
+    public void findOcr(ClassificationPage page) {

        page.getImages().forEach(image -> {
            if (image.getImageType().equals(ImageType.OTHER)) {
@@ -1,4 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.segmentation;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter;

 import java.util.ArrayList;
 import java.util.Collection;
@@ -10,9 +10,9 @@ import org.springframework.stereotype.Service;

 import com.fasterxml.jackson.databind.ObjectMapper;
 import com.iqser.red.service.persistence.service.v1.api.shared.model.dossiertemplate.dossier.file.FileType;
-import com.iqser.red.service.redaction.v1.server.redaction.model.PdfTableCell;
-import com.iqser.red.service.redaction.v1.server.redaction.model.table.TableCells;
-import com.iqser.red.service.redaction.v1.server.redaction.model.table.TableServiceResponse;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.table.PdfTableCell;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.table.TableCells;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.table.TableServiceResponse;
 import com.iqser.red.service.redaction.v1.server.storage.RedactionStorageService;

 import lombok.RequiredArgsConstructor;
@@ -22,7 +22,7 @@ import lombok.extern.slf4j.Slf4j;
@Slf4j
@Service
@RequiredArgsConstructor
-public class TableService {
+public class TableServiceResponseAdapter {

    private final ObjectMapper objectMapper;
    private final RedactionStorageService redactionStorageService;
@@ -1,9 +1,8 @@
-package com.iqser.red.service.redaction.v1.server.redaction.model.image;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.image;

 import java.util.HashMap;
 import java.util.Map;

-
 import lombok.Data;

@Data
@@ -1,5 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.redaction.model.image;
-
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.image;

 import lombok.Data;

@@ -1,5 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.redaction.model.image;
-
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.image;

 import lombok.Data;

@@ -1,5 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.redaction.model.image;
-
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.image;

 import lombok.Data;

@@ -1,5 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.redaction.model.image;
-
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.image;

 import lombok.Data;

@@ -1,5 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.redaction.model.image;
-
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.image;

 import lombok.Data;

@@ -1,13 +1,13 @@
-package com.iqser.red.service.redaction.v1.server.redaction.model.image;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.image;
+
+import java.util.ArrayList;
+import java.util.List;

 import com.fasterxml.jackson.annotation.JsonAlias;
 import com.fasterxml.jackson.annotation.JsonProperty;

 import lombok.Data;

-import java.util.ArrayList;
-import java.util.List;
-
@Data
 public class ImageServiceResponse {

@@ -1,4 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.redaction.model.image;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.image;

 import lombok.Data;

@@ -1,5 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.redaction.model.image;
-
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.image;

 import lombok.Data;

@@ -0,0 +1,10 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.image;
+
+import lombok.Data;
+
+@Data
+public class Probability {
+
+    private boolean unconfident;
+
+}
@@ -1,5 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.redaction.model.table;
-
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.table;

 import lombok.Data;

@@ -1,4 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.redaction.model;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.table;

 import lombok.AllArgsConstructor;
 import lombok.Builder;
@@ -1,5 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.redaction.model.table;
-
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.table;

 import lombok.Data;

@@ -1,9 +1,8 @@
-package com.iqser.red.service.redaction.v1.server.redaction.model.table;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.table;

 import java.util.ArrayList;
 import java.util.List;

-
 import lombok.Data;

@Data
@@ -1,9 +1,8 @@
-package com.iqser.red.service.redaction.v1.server.redaction.model.table;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.table;

 import java.util.ArrayList;
 import java.util.List;

-
 import lombok.Data;

@Data
@@ -1,9 +1,8 @@
-package com.iqser.red.service.redaction.v1.server.tableextraction.model;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model;

 import com.fasterxml.jackson.annotation.JsonIgnore;
 import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.Rectangle;
-import com.iqser.red.service.redaction.v1.server.classification.model.Orientation;
-import com.iqser.red.service.redaction.v1.server.classification.model.TextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.ClassificationTextBlock;

 import lombok.AllArgsConstructor;
 import lombok.Data;
@@ -12,7 +11,7 @@ import lombok.NoArgsConstructor;
@Data
@AllArgsConstructor
@NoArgsConstructor
-public abstract class AbstractTextContainer {
+public abstract class AbstractPageBlock {

    @JsonIgnore
    protected float minX;
@@ -34,13 +33,13 @@ public abstract class AbstractTextContainer {
    public abstract String getText();


-    public boolean containsBlock(TextBlock other) {
+    public boolean containsBlock(ClassificationTextBlock other) {

        return this.minX <= other.getPdfMinX() && this.maxX >= other.getPdfMaxX() && this.minY >= other.getPdfMinY() && this.maxY <= other.getPdfMaxY();
    }


-    public boolean contains(AbstractTextContainer other) {
+    public boolean contains(AbstractPageBlock other) {

        return this.minX <= other.minX && this.maxX >= other.maxX && this.minY >= other.minY && this.maxY <= other.maxY;
    }
@@ -66,4 +65,10 @@ public abstract class AbstractTextContainer {
        return maxX - minX;
    }

+
+    public boolean intersectsY(AbstractPageBlock atc) {
+
+        return this.minY <= atc.getMaxY() && this.maxY >= atc.getMinY();
+    }
+
 }
@@ -1,22 +1,24 @@
-package com.iqser.red.service.redaction.v1.server.classification.model;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model;

 import java.util.ArrayList;
 import java.util.List;

 import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.section.SectionGrid;
-import com.iqser.red.service.redaction.v1.server.redaction.model.DictionaryVersion;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.StringFrequencyCounter;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.UnclassifiedText;
+import com.iqser.red.service.redaction.v1.server.redaction.model.dictionary.DictionaryVersion;

 import lombok.Data;
 import lombok.NoArgsConstructor;

@Data
@NoArgsConstructor
-public class Document {
+public class ClassificationDocument {

-    private List<Page> pages = new ArrayList<>();
-    private List<Paragraph> paragraphs = new ArrayList<>();
-    private List<Header> headers = new ArrayList<>();
-    private List<Footer> footers = new ArrayList<>();
+    private List<ClassificationPage> pages = new ArrayList<>();
+    private List<ClassificationSection> sections = new ArrayList<>();
+    private List<ClassificationHeader> headers = new ArrayList<>();
+    private List<ClassificationFooter> footers = new ArrayList<>();
    private List<UnclassifiedText> unclassifiedTexts = new ArrayList<>();
    private FloatFrequencyCounter textHeightCounter = new FloatFrequencyCounter();
    private FloatFrequencyCounter fontSizeCounter = new FloatFrequencyCounter();
@@ -1,18 +1,19 @@
-package com.iqser.red.service.redaction.v1.server.classification.model;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model;
+
+import java.util.List;

 import com.fasterxml.jackson.annotation.JsonIgnore;
-import com.iqser.red.service.redaction.v1.server.redaction.model.SearchableText;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.ClassificationTextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.SearchableText;

 import lombok.AllArgsConstructor;
 import lombok.Data;

-import java.util.List;
-
@Data
@AllArgsConstructor
-public class Footer {
+public class ClassificationFooter {

-    private List<TextBlock> textBlocks;
+    private List<ClassificationTextBlock> textBlocks;


    @JsonIgnore
@@ -1,18 +1,19 @@
-package com.iqser.red.service.redaction.v1.server.classification.model;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model;
+
+import java.util.List;

 import com.fasterxml.jackson.annotation.JsonIgnore;
-import com.iqser.red.service.redaction.v1.server.redaction.model.SearchableText;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.ClassificationTextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.SearchableText;

 import lombok.AllArgsConstructor;
 import lombok.Data;

-import java.util.List;
-
@Data
@AllArgsConstructor
-public class Header {
+public class ClassificationHeader {

-    private List<TextBlock> textBlocks;
+    private List<ClassificationTextBlock> textBlocks;


    @JsonIgnore
@@ -1,11 +1,11 @@
-package com.iqser.red.service.redaction.v1.server.classification.model;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model;

 import java.util.ArrayList;
 import java.util.List;

 import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.Rectangle;
-import com.iqser.red.service.redaction.v1.server.redaction.model.PdfImage;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.AbstractTextContainer;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.image.ClassifiedImage;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.StringFrequencyCounter;

 import lombok.Data;
 import lombok.NonNull;
@@ -13,12 +13,12 @@ import lombok.RequiredArgsConstructor;

@Data
@RequiredArgsConstructor
-public class Page {
+public class ClassificationPage {

    @NonNull
-    private List<AbstractTextContainer> textBlocks;
+    private List<AbstractPageBlock> textBlocks;

-    private List<PdfImage> images = new ArrayList<>();
+    private List<ClassifiedImage> images = new ArrayList<>();

    private Rectangle bodyTextFrame;

@@ -0,0 +1,65 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model;
+
+import java.util.ArrayList;
+import java.util.List;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.image.ClassifiedImage;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table.TablePageBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.ClassificationTextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.SearchableText;
+
+import lombok.Data;
+import lombok.NoArgsConstructor;
+
+@Data
+@NoArgsConstructor
+public class ClassificationSection implements Comparable {
+
+    private List<AbstractPageBlock> pageBlocks = new ArrayList<>();
+    private List<ClassifiedImage> images = new ArrayList<>();
+    private String headline;
+
+
+    public SearchableText getSearchableText() {
+
+        SearchableText searchableText = new SearchableText();
+        pageBlocks.forEach(block -> {
+            if (block instanceof ClassificationTextBlock) {
+                searchableText.addAll(((ClassificationTextBlock) block).getSequences());
+            }
+        });
+        return searchableText;
+    }
+
+
+    public List<TablePageBlock> getTables() {
+
+        List<TablePageBlock> tables = new ArrayList<>();
+        pageBlocks.forEach(block -> {
+            if (block instanceof TablePageBlock) {
+                tables.add((TablePageBlock) block);
+            }
+        });
+        return tables;
+    }
+
+
+    public List<ClassificationTextBlock> getTextBlocks() {
+
+        List<ClassificationTextBlock> textBlocks = new ArrayList<>();
+        pageBlocks.forEach(block -> {
+            if (block instanceof ClassificationTextBlock) {
+                textBlocks.add((ClassificationTextBlock) block);
+            }
+        });
+        return textBlocks;
+    }
+
+
+    @Override
+    public int compareTo(Object o) {
+
+        return 0;
+    }
+
+}
@@ -1,6 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.classification.model;
-
-import lombok.Getter;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model;

 import java.util.ArrayList;
 import java.util.Collections;
@@ -9,6 +7,8 @@ import java.util.List;
 import java.util.Map;
 import java.util.stream.Collectors;

+import lombok.Getter;
+
 public class FloatFrequencyCounter {

    @Getter
@@ -0,0 +1,8 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model;
+
+public enum Orientation {
+
+    NONE,
+    LEFT,
+    RIGHT
+}
@@ -0,0 +1,25 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.image;
+
+import java.awt.geom.Rectangle2D;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.ImageType;
+
+import lombok.Data;
+import lombok.NonNull;
+import lombok.RequiredArgsConstructor;
+
+@Data
+@RequiredArgsConstructor
+public class ClassifiedImage {
+
+    @NonNull
+    private Rectangle2D position;
+    @NonNull
+    private ImageType imageType;
+    private boolean isAppendedToSection;
+    @NonNull
+    private boolean hasTransparency;
+    @NonNull
+    private int page;
+
+}
@@ -1,25 +1,25 @@
-package com.iqser.red.service.redaction.v1.server.tableextraction.model;
-
-import com.iqser.red.service.redaction.v1.server.classification.model.TextBlock;
-import com.iqser.red.service.redaction.v1.server.parsing.model.TextPositionSequence;
-import com.iqser.red.service.redaction.v1.server.redaction.utils.TextNormalizationUtilities;
-
-import lombok.Data;
-import lombok.EqualsAndHashCode;
-import lombok.NoArgsConstructor;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table;

 import java.awt.geom.Point2D;
 import java.util.ArrayList;
 import java.util.Iterator;
 import java.util.List;

+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.ClassificationTextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.TextPositionSequence;
+import com.iqser.red.service.redaction.v1.server.redaction.utils.TextNormalizationUtilities;
+
+import lombok.Data;
+import lombok.EqualsAndHashCode;
+import lombok.NoArgsConstructor;
+
@SuppressWarnings("serial")
@Data
@EqualsAndHashCode(callSuper = true)
@NoArgsConstructor
 public class Cell extends Rectangle {

-    private List<TextBlock> textBlocks = new ArrayList<>();
+    private List<ClassificationTextBlock> textBlocks = new ArrayList<>();

    private List<Cell> headerCells = new ArrayList<>();

@@ -27,6 +27,8 @@ public class Cell extends Rectangle {

    private static final int MIN_SIZE = 1;

+    private int pageNumber;
+

    public Cell(Point2D topLeft, Point2D bottomRight) {

@@ -34,7 +36,7 @@ public class Cell extends Rectangle {
    }


-    public void addTextBlock(TextBlock textBlock) {
+    public void addTextBlock(ClassificationTextBlock textBlock) {

        textBlocks.add(textBlock);
    }
@@ -45,11 +47,11 @@ public class Cell extends Rectangle {

        StringBuilder sb = new StringBuilder();

-        Iterator<TextBlock> itty = textBlocks.iterator();
+        Iterator<ClassificationTextBlock> itty = textBlocks.iterator();
        TextPositionSequence previous = null;
        while (itty.hasNext()) {

-            TextBlock textBlock = itty.next();
+            ClassificationTextBlock textBlock = itty.next();

            for (TextPositionSequence word : textBlock.getSequences()) {
                if (previous != null) {
@@ -1,4 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.tableextraction.model;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table;

 import lombok.RequiredArgsConstructor;
 import lombok.Value;
@@ -1,10 +1,10 @@
-package com.iqser.red.service.redaction.v1.server.tableextraction.model;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table;
+
+import java.util.List;

 import lombok.Builder;
 import lombok.Data;

-import java.util.List;
-
@Data
@Builder
 public class CleanRulings {
@@ -1,4 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.tableextraction.model;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table;

 import java.awt.geom.Point2D;
 import java.awt.geom.Rectangle2D;
@@ -1,14 +1,20 @@
-package com.iqser.red.service.redaction.v1.server.tableextraction.model;
-
-import com.iqser.red.service.redaction.v1.server.tableextraction.utils.CohenSutherlandClipping;
-import com.iqser.red.service.redaction.v1.server.tableextraction.utils.Utils;
-
-import lombok.extern.slf4j.Slf4j;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table;

 import java.awt.geom.Line2D;
 import java.awt.geom.Point2D;
 import java.awt.geom.Rectangle2D;
-import java.util.*;
+import java.util.ArrayList;
+import java.util.Collections;
+import java.util.Comparator;
+import java.util.Formatter;
+import java.util.List;
+import java.util.Map;
+import java.util.TreeMap;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.utils.CohenSutherlandClipping;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.utils.Utils;
+
+import lombok.extern.slf4j.Slf4j;

@Slf4j
@SuppressWarnings("all")
@@ -1,9 +1,8 @@
-package com.iqser.red.service.redaction.v1.server.tableextraction.model;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table;

 import java.awt.geom.Point2D;
 import java.util.ArrayList;
 import java.util.Collections;
-import java.util.Comparator;
 import java.util.HashSet;
 import java.util.List;
 import java.util.Set;
@@ -12,14 +11,15 @@ import java.util.stream.Collectors;

 import org.apache.commons.collections4.CollectionUtils;

-import com.iqser.red.service.redaction.v1.server.classification.model.TextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.AbstractPageBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.ClassificationTextBlock;

 import lombok.Getter;
 import lombok.Setter;
 import lombok.extern.slf4j.Slf4j;

@Slf4j
-public class Table extends AbstractTextContainer {
+public class TablePageBlock extends AbstractPageBlock {

    private final TreeMap<CellPosition, Cell> cells = new TreeMap<>();

@@ -29,19 +29,17 @@ public class Table extends AbstractTextContainer {
    private String headline;
    private int unrotatedRowCount;
    private int unrotatedColCount;
-    private int rowCount = -1;
-    private int colCount = -1;
    private List<List<Cell>> rows;


-    public Table(List<Cell> cells, Rectangle area, int rotation) {
+    public TablePageBlock(List<Cell> cells, Rectangle area, int rotation) {

        addCells(cells);
        minX = area.getLeft();
        minY = area.getBottom();
        maxX = area.getRight();
        maxY = area.getTop();
-        classification = "Table";
+        classification = "TablePageBlock";
        this.rotation = rotation;

    }
@@ -71,19 +69,13 @@ public class Table extends AbstractTextContainer {

    public int getRowCount() {

-        if (rowCount == -1) {
-            rowCount = getRows().size();
-        }
-        return rowCount;
+        return getRows().size();
    }


    public int getColCount() {

-        if (colCount == -1) {
-            colCount = getRows().stream().mapToInt(List::size).max().orElse(0);
-        }
-        return colCount;
+        return getRows().stream().mapToInt(List::size).max().orElse(0);

    }

@@ -224,7 +216,7 @@ public class Table extends AbstractTextContainer {
     * Calculates the structure of the table. For spanning rows and columns multiple cells with the same values will be inserted.
     *
     * @param cells The found cells
-     * @return Table Structure
+     * @return TablePageBlock Structure
     */
    private List<List<Cell>> calculateStructure(List<Cell> cells) {

@@ -268,7 +260,7 @@ public class Table extends AbstractTextContainer {
                prevX = x;
            }

-            if (prevY != null && prevX != null) {
+            if (prevY != null && prevX != null && !row.isEmpty()) {
                matrix.add(row);
            }
            prevY = y;
@@ -299,7 +291,7 @@ public class Table extends AbstractTextContainer {
                    }
                    if (column != null && column.getTextBlocks() != null) {
                        boolean first = true;
-                        for (TextBlock textBlock : column.getTextBlocks()) {
+                        for (ClassificationTextBlock textBlock : column.getTextBlocks()) {
                            if (!first) {
                                sb.append("\n");
                            }
@@ -331,7 +323,7 @@ public class Table extends AbstractTextContainer {
                    sb.append(i == 0 ? "\n<th>" : "\n<td>");
                    if (column != null && column.getTextBlocks() != null) {
                        boolean first = true;
-                        for (TextBlock textBlock : column.getTextBlocks()) {
+                        for (ClassificationTextBlock textBlock : column.getTextBlocks()) {
                            if (!first) {
                                sb.append("<br />");
                            }
@@ -1,13 +1,11 @@
-package com.iqser.red.service.redaction.v1.server.classification.model;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text;

 import java.util.ArrayList;
 import java.util.List;

 import com.fasterxml.jackson.annotation.JsonIgnore;
-import com.iqser.red.service.redaction.v1.server.parsing.model.TextDirection;
-import com.iqser.red.service.redaction.v1.server.parsing.model.TextPositionSequence;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.AbstractPageBlock;
 import com.iqser.red.service.redaction.v1.server.redaction.utils.TextNormalizationUtilities;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.AbstractTextContainer;

 import lombok.AllArgsConstructor;
 import lombok.Builder;
@@ -18,7 +16,7 @@ import lombok.NoArgsConstructor;
@Builder
@Data
@NoArgsConstructor
-public class TextBlock extends AbstractTextContainer {
+public class ClassificationTextBlock extends AbstractPageBlock {

    @Builder.Default
    private List<TextPositionSequence> sequences = new ArrayList<>();
@@ -26,6 +24,8 @@ public class TextBlock extends AbstractTextContainer {
    @JsonIgnore
    private int rotation;

+    private int indexOnPage;
+
    @JsonIgnore
    private String mostPopularWordFont;

@@ -95,6 +95,7 @@ public class TextBlock extends AbstractTextContainer {
        }
    }

+
    /**
     * Returns the maxX value in pdf coordinate system.
     * Note: This needs to use Pdf Coordinate System where {0,0} rotated with the page rotation.
@@ -174,8 +175,9 @@ public class TextBlock extends AbstractTextContainer {
    }


-    public TextBlock(float minX, float maxX, float minY, float maxY, List<TextPositionSequence> sequences, int rotation) {
+    public ClassificationTextBlock(float minX, float maxX, float minY, float maxY, List<TextPositionSequence> sequences, int rotation, int indexOnPage) {

+        this.indexOnPage = indexOnPage;
        this.minX = minX;
        this.maxX = maxX;
        this.minY = minY;
@@ -185,23 +187,23 @@ public class TextBlock extends AbstractTextContainer {
    }


-    public TextBlock union(TextPositionSequence r) {
+    public ClassificationTextBlock union(TextPositionSequence r) {

-        TextBlock union = this.copy();
+        ClassificationTextBlock union = this.copy();
        union.add(r);
        return union;
    }


-    public TextBlock union(TextBlock r) {
+    public ClassificationTextBlock union(ClassificationTextBlock r) {

-        TextBlock union = this.copy();
+        ClassificationTextBlock union = this.copy();
        union.add(r);
        return union;
    }


-    public void add(TextBlock r) {
+    public void add(ClassificationTextBlock r) {

        if (r.getMinX() < minX) {
            minX = r.getMinX();
@@ -236,9 +238,9 @@ public class TextBlock extends AbstractTextContainer {
    }


-    public TextBlock copy() {
+    public ClassificationTextBlock copy() {

-        return new TextBlock(minX, maxX, minY, maxY, sequences, rotation);
+        return new ClassificationTextBlock(minX, maxX, minY, maxY, sequences, rotation, indexOnPage);
    }


@@ -1,4 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.parsing.model;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text;

 import org.apache.pdfbox.text.TextPosition;
 import org.springframework.beans.BeanUtils;
@@ -0,0 +1,45 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text;
+
+import java.util.ArrayList;
+import java.util.List;
+
+import com.iqser.red.service.redaction.v1.server.redaction.utils.TextNormalizationUtilities;
+
+import lombok.Getter;
+
+public class SearchableText {
+
+    @Getter
+    private final List<TextPositionSequence> sequences = new ArrayList<>();
+
+
+    public void add(TextPositionSequence textPositionSequence) {
+
+        sequences.add(textPositionSequence);
+    }
+
+
+    public void addAll(List<TextPositionSequence> textPositionSequences) {
+
+        sequences.addAll(textPositionSequences);
+    }
+
+
+    @Override
+    public String toString() {
+
+        return buildString(sequences);
+    }
+
+
+    public static String buildString(List<TextPositionSequence> sequences) {
+
+        StringBuilder sb = new StringBuilder();
+        for (TextPositionSequence word : sequences) {
+            sb.append(word.toString());
+            sb.append(' ');
+        }
+        return TextNormalizationUtilities.removeHyphenLineBreaks(sb.toString()).replaceAll("\n", " ").replaceAll(" {2}", " ");
+    }
+
+}
@@ -1,5 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.classification.model;
-
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text;

 import lombok.AllArgsConstructor;
 import lombok.Builder;
@@ -1,4 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.classification.model;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text;

 import java.util.ArrayList;
 import java.util.List;
@@ -1,10 +1,10 @@
-package com.iqser.red.service.redaction.v1.server.classification.model;
-
-import lombok.Getter;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text;

 import java.util.HashMap;
 import java.util.Map;

+import lombok.Getter;
+
 public class StringFrequencyCounter {

    @Getter
@@ -1,4 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.parsing.model;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text;

 import java.util.Objects;

@@ -1,4 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.parsing.model;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text;

 import java.awt.geom.AffineTransform;
 import java.awt.geom.Point2D;
@@ -1,18 +1,17 @@
-package com.iqser.red.service.redaction.v1.server.classification.model;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text;
+
+import java.util.List;

 import com.fasterxml.jackson.annotation.JsonIgnore;
-import com.iqser.red.service.redaction.v1.server.redaction.model.SearchableText;

 import lombok.AllArgsConstructor;
 import lombok.Data;

-import java.util.List;
-
@Data
@AllArgsConstructor
 public class UnclassifiedText {

-    private List<TextBlock> textBlocks;
+    private List<ClassificationTextBlock> textBlocks;


    @JsonIgnore
@@ -14,34 +14,18 @@
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */
-package com.iqser.red.service.redaction.v1.server.parsing;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.parsing;

-import java.io.InputStream;
 import java.io.IOException;
+import java.io.InputStream;
 import java.util.Map;
 import java.util.WeakHashMap;

 import org.apache.commons.logging.Log;
 import org.apache.commons.logging.LogFactory;
-
 import org.apache.fontbox.ttf.TrueTypeFont;
 import org.apache.fontbox.util.BoundingBox;
-
 import org.apache.pdfbox.contentstream.PDFStreamEngine;
-import org.apache.pdfbox.pdmodel.PDPage;
-import org.apache.pdfbox.pdmodel.font.encoding.GlyphList;
-import org.apache.pdfbox.pdmodel.common.PDRectangle;
-import org.apache.pdfbox.pdmodel.font.PDCIDFont;
-import org.apache.pdfbox.pdmodel.font.PDCIDFontType2;
-import org.apache.pdfbox.pdmodel.font.PDFont;
-import org.apache.pdfbox.pdmodel.font.PDSimpleFont;
-import org.apache.pdfbox.pdmodel.font.PDTrueTypeFont;
-import org.apache.pdfbox.pdmodel.font.PDType0Font;
-import org.apache.pdfbox.pdmodel.font.PDType3Font;
-import org.apache.pdfbox.pdmodel.graphics.state.PDGraphicsState;
-import org.apache.pdfbox.text.TextPosition;
-import org.apache.pdfbox.util.Matrix;
-import org.apache.pdfbox.util.Vector;
 import org.apache.pdfbox.contentstream.operator.DrawObject;
 import org.apache.pdfbox.contentstream.operator.state.Concatenate;
 import org.apache.pdfbox.contentstream.operator.state.Restore;
@@ -50,22 +34,36 @@ import org.apache.pdfbox.contentstream.operator.state.SetGraphicsStateParameters
 import org.apache.pdfbox.contentstream.operator.state.SetMatrix;
 import org.apache.pdfbox.contentstream.operator.text.BeginText;
 import org.apache.pdfbox.contentstream.operator.text.EndText;
-import org.apache.pdfbox.contentstream.operator.text.SetFontAndSize;
-import org.apache.pdfbox.contentstream.operator.text.SetTextHorizontalScaling;
-import org.apache.pdfbox.contentstream.operator.text.ShowTextAdjusted;
-import org.apache.pdfbox.contentstream.operator.text.ShowTextLine;
-import org.apache.pdfbox.contentstream.operator.text.ShowTextLineAndSpace;
 import org.apache.pdfbox.contentstream.operator.text.MoveText;
 import org.apache.pdfbox.contentstream.operator.text.MoveTextSetLeading;
 import org.apache.pdfbox.contentstream.operator.text.NextLine;
 import org.apache.pdfbox.contentstream.operator.text.SetCharSpacing;
+import org.apache.pdfbox.contentstream.operator.text.SetFontAndSize;
+import org.apache.pdfbox.contentstream.operator.text.SetTextHorizontalScaling;
 import org.apache.pdfbox.contentstream.operator.text.SetTextLeading;
 import org.apache.pdfbox.contentstream.operator.text.SetTextRenderingMode;
 import org.apache.pdfbox.contentstream.operator.text.SetTextRise;
 import org.apache.pdfbox.contentstream.operator.text.SetWordSpacing;
 import org.apache.pdfbox.contentstream.operator.text.ShowText;
+import org.apache.pdfbox.contentstream.operator.text.ShowTextAdjusted;
+import org.apache.pdfbox.contentstream.operator.text.ShowTextLine;
+import org.apache.pdfbox.contentstream.operator.text.ShowTextLineAndSpace;
 import org.apache.pdfbox.cos.COSDictionary;
+import org.apache.pdfbox.pdmodel.PDPage;
+import org.apache.pdfbox.pdmodel.common.PDRectangle;
+import org.apache.pdfbox.pdmodel.font.PDCIDFont;
+import org.apache.pdfbox.pdmodel.font.PDCIDFontType2;
+import org.apache.pdfbox.pdmodel.font.PDFont;
 import org.apache.pdfbox.pdmodel.font.PDFontDescriptor;
+import org.apache.pdfbox.pdmodel.font.PDSimpleFont;
+import org.apache.pdfbox.pdmodel.font.PDTrueTypeFont;
+import org.apache.pdfbox.pdmodel.font.PDType0Font;
+import org.apache.pdfbox.pdmodel.font.PDType3Font;
+import org.apache.pdfbox.pdmodel.font.encoding.GlyphList;
+import org.apache.pdfbox.pdmodel.graphics.state.PDGraphicsState;
+import org.apache.pdfbox.text.TextPosition;
+import org.apache.pdfbox.util.Matrix;
+import org.apache.pdfbox.util.Vector;

 /**
 * LEGACY text calculations which are known to be incorrect but are depended on by PDFTextStripper.
@@ -1,17 +1,17 @@
-package com.iqser.red.service.redaction.v1.server.parsing;
-
-import com.iqser.red.service.redaction.v1.server.parsing.model.TextPositionSequence;
-
-import lombok.Getter;
-import lombok.Setter;
-
-import org.apache.pdfbox.text.PDFTextStripperByArea;
-import org.apache.pdfbox.text.TextPosition;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.parsing;

 import java.io.IOException;
 import java.util.ArrayList;
 import java.util.List;

+import org.apache.pdfbox.text.PDFTextStripperByArea;
+import org.apache.pdfbox.text.TextPosition;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.TextPositionSequence;
+
+import lombok.Getter;
+import lombok.Setter;
+
 public class PDFAreaTextStripper extends PDFTextStripperByArea {

    @Getter
@@ -1,17 +1,32 @@
-package com.iqser.red.service.redaction.v1.server.parsing;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.parsing;

-import com.iqser.red.service.redaction.v1.server.parsing.model.RedTextPosition;
-import com.iqser.red.service.redaction.v1.server.parsing.model.TextPositionSequence;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.Ruling;
-
-import lombok.Getter;
-import lombok.Setter;
-import lombok.extern.slf4j.Slf4j;
+import java.awt.geom.Point2D;
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.Comparator;
+import java.util.List;

 import org.apache.pdfbox.contentstream.operator.Operator;
 import org.apache.pdfbox.contentstream.operator.OperatorName;
-import org.apache.pdfbox.contentstream.operator.color.*;
-import org.apache.pdfbox.contentstream.operator.state.*;
+import org.apache.pdfbox.contentstream.operator.color.SetNonStrokingColor;
+import org.apache.pdfbox.contentstream.operator.color.SetNonStrokingColorN;
+import org.apache.pdfbox.contentstream.operator.color.SetNonStrokingColorSpace;
+import org.apache.pdfbox.contentstream.operator.color.SetNonStrokingDeviceCMYKColor;
+import org.apache.pdfbox.contentstream.operator.color.SetNonStrokingDeviceGrayColor;
+import org.apache.pdfbox.contentstream.operator.color.SetNonStrokingDeviceRGBColor;
+import org.apache.pdfbox.contentstream.operator.color.SetStrokingColor;
+import org.apache.pdfbox.contentstream.operator.color.SetStrokingColorN;
+import org.apache.pdfbox.contentstream.operator.color.SetStrokingColorSpace;
+import org.apache.pdfbox.contentstream.operator.color.SetStrokingDeviceCMYKColor;
+import org.apache.pdfbox.contentstream.operator.color.SetStrokingDeviceGrayColor;
+import org.apache.pdfbox.contentstream.operator.color.SetStrokingDeviceRGBColor;
+import org.apache.pdfbox.contentstream.operator.state.SetFlatness;
+import org.apache.pdfbox.contentstream.operator.state.SetLineCapStyle;
+import org.apache.pdfbox.contentstream.operator.state.SetLineDashPattern;
+import org.apache.pdfbox.contentstream.operator.state.SetLineJoinStyle;
+import org.apache.pdfbox.contentstream.operator.state.SetLineMiterLimit;
+import org.apache.pdfbox.contentstream.operator.state.SetLineWidth;
+import org.apache.pdfbox.contentstream.operator.state.SetRenderingIntent;
 import org.apache.pdfbox.contentstream.operator.text.SetFontAndSize;
 import org.apache.pdfbox.cos.COSBase;
 import org.apache.pdfbox.cos.COSNumber;
@@ -19,11 +34,13 @@ import org.apache.pdfbox.pdmodel.PDDocument;
 import org.apache.pdfbox.pdmodel.PDPage;
 import org.apache.pdfbox.text.TextPosition;

-import java.awt.geom.Point2D;
-import java.io.IOException;
-import java.util.ArrayList;
-import java.util.Comparator;
-import java.util.List;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table.Ruling;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.RedTextPosition;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.TextPositionSequence;
+
+import lombok.Getter;
+import lombok.Setter;
+import lombok.extern.slf4j.Slf4j;

@Slf4j
 public class PDFLinesTextStripper extends PDFTextStripper {
@@ -264,8 +281,8 @@ public class PDFLinesTextStripper extends PDFTextStripper {
                    // Remove false sequence ends (whitespaces)
                    if (previous != null && sublist.get(0).getYDirAdj() == previous.getYDirAdj() && sublist.get(0)
                            .getXDirAdj() - (previous.getXDirAdj() + previous.getWidthDirAdj()) < 0.01) {
-                        for (TextPosition t : sublist) {
-                            textPositionSequences.get(textPositionSequences.size() - 1).add(t);
+                        for (TextPosition textPosition : sublist) {
+                            textPositionSequences.get(textPositionSequences.size() - 1).add(textPosition);
                        }
                    } else {
                        textPositionSequences.add(new TextPositionSequence(sublist, pageNumber));
@@ -14,7 +14,7 @@
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */
-package com.iqser.red.service.redaction.v1.server.parsing;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.parsing;

 import java.io.BufferedInputStream;
 import java.io.IOException;
@@ -1,4 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.classification.service;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.service;

 import static java.util.stream.Collectors.toSet;

@@ -9,15 +9,15 @@ import java.util.List;

 import org.springframework.stereotype.Service;

-import com.iqser.red.service.redaction.v1.server.classification.model.FloatFrequencyCounter;
-import com.iqser.red.service.redaction.v1.server.classification.model.Orientation;
-import com.iqser.red.service.redaction.v1.server.classification.model.Page;
-import com.iqser.red.service.redaction.v1.server.classification.model.StringFrequencyCounter;
-import com.iqser.red.service.redaction.v1.server.classification.model.TextBlock;
-import com.iqser.red.service.redaction.v1.server.classification.utils.RulingTextDirAdjustUtil;
-import com.iqser.red.service.redaction.v1.server.parsing.model.TextPositionSequence;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.AbstractTextContainer;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.Ruling;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.AbstractPageBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.ClassificationPage;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.FloatFrequencyCounter;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.Orientation;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table.Ruling;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.ClassificationTextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.StringFrequencyCounter;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.TextPositionSequence;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.utils.RulingTextDirAdjustUtil;

@Service
@SuppressWarnings("all")
@@ -29,16 +29,18 @@ public class BlockificationService {
    /**
     * This method is building blocks by expanding the minX/maxX and minY/maxY value on each word that is not split by the conditions.
     * This method must use text direction adjusted postions (DirAdj). Where {0,0} is on the upper left. Never try to change this!
-     * Rulings (Table lines) must be adjusted to the text directions as well, when checking if a block is split by a ruling.
-     * @param textPositions The words of a page.
+     * Rulings (TablePageBlock lines) must be adjusted to the text directions as well, when checking if a block is split by a ruling.
+     *
+     * @param textPositions         The words of a page.
     * @param horizontalRulingLines Horizontal table lines.
-     * @param verticalRulingLines Vertical table lines.
+     * @param verticalRulingLines   Vertical table lines.
     * @return Page object that contains the Textblock and text statistics.
     */
-    public Page blockify(List<TextPositionSequence> textPositions, List<Ruling> horizontalRulingLines, List<Ruling> verticalRulingLines) {
+    public ClassificationPage blockify(List<TextPositionSequence> textPositions, List<Ruling> horizontalRulingLines, List<Ruling> verticalRulingLines) {

+        int indexOnPage = 0;
        List<TextPositionSequence> chunkWords = new ArrayList<>();
-        List<AbstractTextContainer> chunkBlockList1 = new ArrayList<>();
+        List<AbstractPageBlock> chunkBlockList1 = new ArrayList<>();

        float minX = 1000, maxX = 0, minY = 1000, maxY = 0;
        TextPositionSequence prev = null;
@@ -62,7 +64,9 @@ public class BlockificationService {
                    prevOrientation = chunkBlockList1.get(chunkBlockList1.size() - 1).getOrientation();
                }

-                TextBlock cb1 = buildTextBlock(chunkWords);
+                ClassificationTextBlock cb1 = buildTextBlock(chunkWords, indexOnPage);
+                indexOnPage++;
+
                chunkBlockList1.add(cb1);
                chunkWords = new ArrayList<>();

@@ -102,17 +106,17 @@ public class BlockificationService {
            }
        }

-        TextBlock cb1 = buildTextBlock(chunkWords);
+        ClassificationTextBlock cb1 = buildTextBlock(chunkWords, indexOnPage);
        if (cb1 != null) {
            chunkBlockList1.add(cb1);
        }

-        Iterator<AbstractTextContainer> itty = chunkBlockList1.iterator();
+        Iterator<AbstractPageBlock> itty = chunkBlockList1.iterator();

-        TextBlock previousLeft = null;
-        TextBlock previousRight = null;
+        ClassificationTextBlock previousLeft = null;
+        ClassificationTextBlock previousRight = null;
        while (itty.hasNext()) {
-            TextBlock block = (TextBlock) itty.next();
+            ClassificationTextBlock block = (ClassificationTextBlock) itty.next();

            if (previousLeft != null && block.getOrientation().equals(Orientation.LEFT)) {
                if (previousLeft.getMinY() > block.getMinY() && block.getMaxY() + block.getMostPopularWordHeight() > previousLeft.getMinY()) {
@@ -138,9 +142,9 @@ public class BlockificationService {
        }

        itty = chunkBlockList1.iterator();
-        TextBlock previous = null;
+        ClassificationTextBlock previous = null;
        while (itty.hasNext()) {
-            TextBlock block = (TextBlock) itty.next();
+            ClassificationTextBlock block = (ClassificationTextBlock) itty.next();

            if (previous != null && previous.getOrientation().equals(Orientation.LEFT) && block.getOrientation().equals(Orientation.LEFT) && equalsWithThreshold(block.getMaxY(),
                    previous.getMaxY()) || previous != null && previous.getOrientation().equals(Orientation.LEFT) && block.getOrientation()
@@ -153,7 +157,7 @@ public class BlockificationService {
            previous = block;
        }

-        return new Page(chunkBlockList1);
+        return new ClassificationPage(chunkBlockList1);
    }


@@ -163,9 +167,9 @@ public class BlockificationService {
    }


-    private TextBlock buildTextBlock(List<TextPositionSequence> wordBlockList) {
+    private ClassificationTextBlock buildTextBlock(List<TextPositionSequence> wordBlockList, int indexOnPage) {

-        TextBlock textBlock = null;
+        ClassificationTextBlock textBlock = null;

        FloatFrequencyCounter lineHeightFrequencyCounter = new FloatFrequencyCounter();
        FloatFrequencyCounter fontSizeFrequencyCounter = new FloatFrequencyCounter();
@@ -182,9 +186,15 @@ public class BlockificationService {
            styleFrequencyCounter.add(wordBlock.getFontStyle());

            if (textBlock == null) {
-                textBlock = new TextBlock(wordBlock.getMinXDirAdj(), wordBlock.getMaxXDirAdj(), wordBlock.getMinYDirAdj(), wordBlock.getMaxYDirAdj(), wordBlockList, wordBlock.getRotation());
+                textBlock = new ClassificationTextBlock(wordBlock.getMinXDirAdj(),
+                        wordBlock.getMaxXDirAdj(),
+                        wordBlock.getMinYDirAdj(),
+                        wordBlock.getMaxYDirAdj(),
+                        wordBlockList,
+                        wordBlock.getRotation(),
+                        indexOnPage);
            } else {
-                TextBlock spatialEntity = textBlock.union(wordBlock);
+                ClassificationTextBlock spatialEntity = textBlock.union(wordBlock);
                textBlock.resize(spatialEntity.getMinX(), spatialEntity.getMinY(), spatialEntity.getWidth(), spatialEntity.getHeight());
            }
        }
@@ -213,10 +223,38 @@ public class BlockificationService {
                                    List<Ruling> horizontalRulingLines,
                                    List<Ruling> verticalRulingLines) {

-        return isSplitByRuling(maxX, minY, word.getMinXDirAdj(), word.getMinYDirAdj(), verticalRulingLines, word.getDir().getDegrees(), word.getPageWidth(), word.getPageHeight()) //
-                || isSplitByRuling(minX, minY, word.getMinXDirAdj(), word.getMaxYDirAdj(), horizontalRulingLines, word.getDir().getDegrees(), word.getPageWidth(), word.getPageHeight()) //
-                || isSplitByRuling(maxX, minY, word.getMinXDirAdj(), word.getMinYDirAdj(), horizontalRulingLines, word.getDir().getDegrees(), word.getPageWidth(), word.getPageHeight()) //
-                || isSplitByRuling(minX, minY, word.getMinXDirAdj(), word.getMaxYDirAdj(), verticalRulingLines, word.getDir().getDegrees(), word.getPageWidth(), word.getPageHeight()); //
+        return isSplitByRuling(maxX,
+                minY,
+                word.getMinXDirAdj(),
+                word.getMinYDirAdj(),
+                verticalRulingLines,
+                word.getDir().getDegrees(),
+                word.getPageWidth(),
+                word.getPageHeight()) //
+                || isSplitByRuling(minX,
+                minY,
+                word.getMinXDirAdj(),
+                word.getMaxYDirAdj(),
+                horizontalRulingLines,
+                word.getDir().getDegrees(),
+                word.getPageWidth(),
+                word.getPageHeight()) //
+                || isSplitByRuling(maxX,
+                minY,
+                word.getMinXDirAdj(),
+                word.getMinYDirAdj(),
+                horizontalRulingLines,
+                word.getDir().getDegrees(),
+                word.getPageWidth(),
+                word.getPageHeight()) //
+                || isSplitByRuling(minX,
+                minY,
+                word.getMinXDirAdj(),
+                word.getMaxYDirAdj(),
+                verticalRulingLines,
+                word.getDir().getDegrees(),
+                word.getPageWidth(),
+                word.getPageHeight()); //
    }


@@ -1,4 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.classification.service;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.service;

 import java.util.List;

@@ -6,13 +6,13 @@ import org.springframework.stereotype.Service;

 import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.Point;
 import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.Rectangle;
-import com.iqser.red.service.redaction.v1.server.classification.model.FloatFrequencyCounter;
-import com.iqser.red.service.redaction.v1.server.classification.model.Page;
-import com.iqser.red.service.redaction.v1.server.classification.model.TextBlock;
-import com.iqser.red.service.redaction.v1.server.classification.utils.PositionUtils;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.AbstractTextContainer;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.Cell;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.Table;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.AbstractPageBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.ClassificationPage;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.FloatFrequencyCounter;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table.Cell;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table.TablePageBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.ClassificationTextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.utils.PositionUtils;

@Service
 public class BodyTextFrameService {
@@ -30,7 +30,7 @@ public class BodyTextFrameService {
     * @param bodyTextFrame          frame that contains the main text on portrait pages
     * @param landscapeBodyTextFrame frame that contains the main text on landscape pages
     */
-    public void setBodyTextFrameAdjustedToPage(Page page, Rectangle bodyTextFrame, Rectangle landscapeBodyTextFrame) {
+    public void setBodyTextFrameAdjustedToPage(ClassificationPage page, Rectangle bodyTextFrame, Rectangle landscapeBodyTextFrame) {

        Rectangle textFrame = page.isLandscape() ? landscapeBodyTextFrame : bodyTextFrame;

@@ -65,20 +65,20 @@ public class BodyTextFrameService {
     * @param landscape               Calculate for landscape or portrait
     * @return Rectangle of the text frame
     */
-    public Rectangle calculateBodyTextFrame(List<Page> pages, FloatFrequencyCounter documentFontSizeCounter, boolean landscape) {
+    public Rectangle calculateBodyTextFrame(List<ClassificationPage> pages, FloatFrequencyCounter documentFontSizeCounter, boolean landscape) {

        BodyTextFrameExpansionsRectangle expansionsRectangle = new BodyTextFrameExpansionsRectangle();

-        for (Page page : pages) {
+        for (ClassificationPage page : pages) {

            if (page.getTextBlocks().isEmpty() || landscape != page.isLandscape()) {
                continue;
            }

-            for (AbstractTextContainer container : page.getTextBlocks()) {
+            for (AbstractPageBlock container : page.getTextBlocks()) {

-                if (container instanceof TextBlock) {
-                    TextBlock textBlock = (TextBlock) container;
+                if (container instanceof ClassificationTextBlock) {
+                    ClassificationTextBlock textBlock = (ClassificationTextBlock) container;
                    if (textBlock.getMostPopularWordFont() == null || textBlock.getMostPopularWordStyle() == null) {
                        continue;
                    }
@@ -94,15 +94,15 @@ public class BodyTextFrameService {
                    }
                }

-                if (container instanceof Table) {
-                    Table table = (Table) container;
+                if (container instanceof TablePageBlock) {
+                    TablePageBlock table = (TablePageBlock) container;
                    for (List<Cell> row : table.getRows()) {
                        for (Cell cell : row) {

                            if (cell == null || cell.getTextBlocks() == null) {
                                continue;
                            }
-                            for (TextBlock textBlock : cell.getTextBlocks()) {
+                            for (ClassificationTextBlock textBlock : cell.getTextBlocks()) {
                                expandRectangle(textBlock, page, expansionsRectangle);
                            }
                        }
@@ -117,7 +117,7 @@ public class BodyTextFrameService {
    }


-    private void expandRectangle(TextBlock textBlock, Page page, BodyTextFrameExpansionsRectangle expansionsRectangle) {
+    private void expandRectangle(ClassificationTextBlock textBlock, ClassificationPage page, BodyTextFrameExpansionsRectangle expansionsRectangle) {

        if (page.getPageWidth() > page.getPageHeight() && page.getRotation() != 0) {
            if (textBlock.getPdfMinY() < expansionsRectangle.minX) {
@@ -1,4 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.classification.service;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.service;

 import java.util.List;
 import java.util.regex.Pattern;
@@ -6,11 +6,11 @@ import java.util.regex.Pattern;
 import org.springframework.stereotype.Service;

 import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.Rectangle;
-import com.iqser.red.service.redaction.v1.server.classification.model.Document;
-import com.iqser.red.service.redaction.v1.server.classification.model.Page;
-import com.iqser.red.service.redaction.v1.server.classification.model.TextBlock;
-import com.iqser.red.service.redaction.v1.server.classification.utils.PositionUtils;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.AbstractTextContainer;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.AbstractPageBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.ClassificationDocument;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.ClassificationPage;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.ClassificationTextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.utils.PositionUtils;

 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;
@@ -23,7 +23,7 @@ public class ClassificationService {
    private final BodyTextFrameService bodyTextFrameService;


-    public void classifyDocument(Document document) {
+    public void classifyDocument(ClassificationDocument document) {

        Rectangle bodyTextFrame = bodyTextFrameService.calculateBodyTextFrame(document.getPages(), document.getFontSizeCounter(), false);
        Rectangle landscapeBodyTextFrame = bodyTextFrameService.calculateBodyTextFrame(document.getPages(), document.getFontSizeCounter(), true);
@@ -31,24 +31,24 @@ public class ClassificationService {

        log.debug("Document FontSize counters are: {}", document.getFontSizeCounter().getCountPerValue());

-        for (Page page : document.getPages()) {
+        for (ClassificationPage page : document.getPages()) {
            bodyTextFrameService.setBodyTextFrameAdjustedToPage(page, bodyTextFrame, landscapeBodyTextFrame);
            classifyPage(page, document, headlineFontSizes);
        }
    }


-    public void classifyPage(Page page, Document document, List<Float> headlineFontSizes) {
+    public void classifyPage(ClassificationPage page, ClassificationDocument document, List<Float> headlineFontSizes) {

-        for (AbstractTextContainer textBlock : page.getTextBlocks()) {
-            if (textBlock instanceof TextBlock) {
-                classifyBlock((TextBlock) textBlock, page, document, headlineFontSizes);
+        for (AbstractPageBlock textBlock : page.getTextBlocks()) {
+            if (textBlock instanceof ClassificationTextBlock) {
+                classifyBlock((ClassificationTextBlock) textBlock, page, document, headlineFontSizes);
            }
        }
    }


-    public void classifyBlock(TextBlock textBlock, Page page, Document document, List<Float> headlineFontSizes) {
+    public void classifyBlock(ClassificationTextBlock textBlock, ClassificationPage page, ClassificationDocument document, List<Float> headlineFontSizes) {

        var bodyTextFrame = page.getBodyTextFrame();

@@ -84,7 +84,7 @@ public class ClassificationService {
                    document.setHeadlines(true);
                }
            }
-        } else if (!textBlock.getText().startsWith("Table ") && !textBlock.getText().startsWith("Figure ") && PositionUtils.isWithinBodyTextFrame(bodyTextFrame,
+        } else if (!textBlock.getText().startsWith("TablePageBlock ") && !textBlock.getText().startsWith("Figure ") && PositionUtils.isWithinBodyTextFrame(bodyTextFrame,
                textBlock) && textBlock.getMostPopularWordStyle().equals("bold") && !document.getFontStyleCounter()
                .getMostPopular()
                .equals("bold") && PositionUtils.getApproxLineCount(textBlock) < 2.9 && textBlock.getSequences()
@@ -1,4 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.segmentation;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.service;

 import java.io.File;
 import java.io.FileOutputStream;
@@ -16,21 +16,19 @@ import org.apache.pdfbox.pdmodel.PDPage;
 import org.apache.pdfbox.pdmodel.common.PDRectangle;
 import org.springframework.stereotype.Service;

-import com.iqser.red.service.redaction.v1.server.classification.model.Document;
-import com.iqser.red.service.redaction.v1.server.classification.model.Page;
-import com.iqser.red.service.redaction.v1.server.classification.model.TextBlock;
-import com.iqser.red.service.redaction.v1.server.classification.service.BlockificationService;
-import com.iqser.red.service.redaction.v1.server.classification.service.ClassificationService;
-import com.iqser.red.service.redaction.v1.server.parsing.PDFLinesTextStripper;
-import com.iqser.red.service.redaction.v1.server.parsing.model.TextPositionSequence;
-import com.iqser.red.service.redaction.v1.server.redaction.model.PdfImage;
-import com.iqser.red.service.redaction.v1.server.redaction.model.PdfTableCell;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.ImageServiceResponseAdapter;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.TableServiceResponseAdapter;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.table.PdfTableCell;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.AbstractPageBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.ClassificationDocument;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.ClassificationPage;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.image.ClassifiedImage;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table.CleanRulings;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.ClassificationTextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.TextPositionSequence;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.parsing.PDFLinesTextStripper;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.utils.FileUtils;
 import com.iqser.red.service.redaction.v1.server.settings.RedactionServiceSettings;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.AbstractTextContainer;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.CleanRulings;
-import com.iqser.red.service.redaction.v1.server.tableextraction.service.RulingCleaningService;
-import com.iqser.red.service.redaction.v1.server.tableextraction.service.TableExtractionService;
-import com.iqser.red.service.redaction.v1.server.tableextraction.utils.FileUtils;

 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;
@@ -46,18 +44,21 @@ public class PdfSegmentationService {
    private final BlockificationService blockificationService;
    private final ClassificationService classificationService;
    private final SectionsBuilderService sectionsBuilderService;
-    private final ImageService imageService;
-    private final TableService tableService;
+    private final ImageServiceResponseAdapter imageServiceResponseAdapter;
+    private final TableServiceResponseAdapter tableServiceResponseAdapter;


-    public Document parseDocument(String dossierId, String fileId, InputStream documentInputStream, Map<Integer, List<PdfImage>> pdfImages) throws IOException {
+    public ClassificationDocument parseDocument(String dossierId,
+                                                String fileId,
+                                                InputStream documentInputStream,
+                                                Map<Integer, List<ClassifiedImage>> pdfImages) throws IOException {

        PDDocument pdDocument = null;
        File tempFile = null;
        try {
            Map<Integer, List<PdfTableCell>> pdfTableCells = new HashMap<>();
            if (redactionServiceSettings.isCvTableParsingEnabled()) {
-                pdfTableCells = tableService.convertTables(dossierId, fileId);
+                pdfTableCells = tableServiceResponseAdapter.convertTables(dossierId, fileId);
            }

            tempFile = FileUtils.createTempFile("document", ".pdf");
@@ -65,8 +66,8 @@ public class PdfSegmentationService {
                IOUtils.copy(documentInputStream, fos);

                // initialize required variables
-                Document document = new Document();
-                List<Page> pages = new ArrayList<>();
+                ClassificationDocument document = new ClassificationDocument();
+                List<ClassificationPage> pages = new ArrayList<>();

                pdDocument = PDDocument.load(tempFile, MemoryUsageSetting.setupMixed(67108864L));
                pdDocument.setAllSecurityToBeRemoved(true);
@@ -94,12 +95,12 @@ public class PdfSegmentationService {
    }


-    private void processPage(Map<Integer, List<PdfImage>> pdfImages,
-                           PDDocument pdDocument,
-                           Map<Integer, List<PdfTableCell>> pdfTableCells,
-                           Document document,
-                           List<Page> pages,
-                           int pageNumber) throws IOException {
+    private void processPage(Map<Integer, List<ClassifiedImage>> pdfImages,
+                             PDDocument pdDocument,
+                             Map<Integer, List<PdfTableCell>> pdfTableCells,
+                             ClassificationDocument document,
+                             List<ClassificationPage> pages,
+                             int pageNumber) throws IOException {

        PDFLinesTextStripper stripper = new PDFLinesTextStripper();
        PDPage pdPage = pdDocument.getPage(pageNumber - 1);
@@ -119,7 +120,7 @@ public class PdfSegmentationService {
                stripper.getRulings(),
                stripper.getMinCharWidth(),
                stripper.getMaxCharHeight());
-        Page page = blockificationService.blockify(stripper.getTextPositionSequences(), cleanRulings.getHorizontal(), cleanRulings.getVertical());
+        ClassificationPage page = blockificationService.blockify(stripper.getTextPositionSequences(), cleanRulings.getHorizontal(), cleanRulings.getVertical());

        page.setRotation(rotation);
        page.setLandscape(isLandscape);
@@ -130,7 +131,7 @@ public class PdfSegmentationService {
        // If images is ocr needs to be calculated before textBlocks are moved into tables, otherwise findOcr algorithm needs to be adopted.
        if (pdfImages != null && pdfImages.containsKey(pageNumber)) {
            page.setImages(pdfImages.get(pageNumber));
-            imageService.findOcr(page);
+            imageServiceResponseAdapter.findOcr(page);
        }

        tableExtractionService.extractTables(cleanRulings, page);
@@ -141,7 +142,7 @@ public class PdfSegmentationService {
    }


-    private void increaseDocumentStatistics(Page page, Document document) {
+    private void increaseDocumentStatistics(ClassificationPage page, ClassificationDocument document) {

        if (!page.isLandscape()) {
            document.getFontSizeCounter().addAll(page.getFontSizeCounter().getCountPerValue());
@@ -152,15 +153,15 @@ public class PdfSegmentationService {
    }


-    private void buildPageStatistics(Page page) {
+    private void buildPageStatistics(ClassificationPage page) {

        // Collect all statistics for the page, except from blocks inside tables, as tables will always be added to BodyTextFrame.
-        for (AbstractTextContainer textBlock : page.getTextBlocks()) {
-            if (textBlock instanceof TextBlock) {
-                if (((TextBlock) textBlock).getSequences() == null) {
+        for (AbstractPageBlock textBlock : page.getTextBlocks()) {
+            if (textBlock instanceof ClassificationTextBlock) {
+                if (((ClassificationTextBlock) textBlock).getSequences() == null) {
                    continue;
                }
-                for (TextPositionSequence word : ((TextBlock) textBlock).getSequences()) {
+                for (TextPositionSequence word : ((ClassificationTextBlock) textBlock).getSequences()) {
                    page.getTextHeightCounter().add(word.getTextHeight());
                    page.getFontCounter().add(word.getFont());
                    page.getFontSizeCounter().add(word.getFontSize());
@@ -1,4 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.tableextraction.service;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.service;

 import java.awt.geom.Line2D;
 import java.awt.geom.Point2D;
@@ -12,11 +12,11 @@ import java.util.Map;

 import org.springframework.stereotype.Service;

-import com.iqser.red.service.redaction.v1.server.redaction.model.PdfTableCell;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.adapter.table.PdfTableCell;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table.CleanRulings;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table.Ruling;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.utils.Utils;
 import com.iqser.red.service.redaction.v1.server.settings.RedactionServiceSettings;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.CleanRulings;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.Ruling;
-import com.iqser.red.service.redaction.v1.server.tableextraction.utils.Utils;

 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;
@@ -1,9 +1,8 @@
-package com.iqser.red.service.redaction.v1.server.segmentation;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.service;

 import java.util.ArrayList;
 import java.util.Collections;
 import java.util.HashMap;
-import java.util.Iterator;
 import java.util.List;
 import java.util.Map;
 import java.util.stream.Collectors;
@@ -11,17 +10,17 @@ import java.util.stream.Collectors;
 import org.apache.commons.collections4.CollectionUtils;
 import org.springframework.stereotype.Service;

-import com.iqser.red.service.redaction.v1.server.classification.model.Document;
-import com.iqser.red.service.redaction.v1.server.classification.model.Footer;
-import com.iqser.red.service.redaction.v1.server.classification.model.Header;
-import com.iqser.red.service.redaction.v1.server.classification.model.Page;
-import com.iqser.red.service.redaction.v1.server.classification.model.Paragraph;
-import com.iqser.red.service.redaction.v1.server.classification.model.TextBlock;
-import com.iqser.red.service.redaction.v1.server.classification.model.UnclassifiedText;
-import com.iqser.red.service.redaction.v1.server.redaction.model.PdfImage;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.AbstractTextContainer;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.Cell;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.Table;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.AbstractPageBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.ClassificationDocument;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.ClassificationFooter;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.ClassificationHeader;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.ClassificationPage;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.ClassificationSection;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.image.ClassifiedImage;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table.Cell;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table.TablePageBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.ClassificationTextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.UnclassifiedText;

 import lombok.extern.slf4j.Slf4j;

@@ -29,23 +28,23 @@ import lombok.extern.slf4j.Slf4j;
@Service
 public class SectionsBuilderService {

-    public void buildSections(Document document) {
+    public void buildSections(ClassificationDocument document) {

-        List<AbstractTextContainer> chunkWords = new ArrayList<>();
-        List<Paragraph> chunkBlockList = new ArrayList<>();
-        List<Header> headers = new ArrayList<>();
-        List<Footer> footers = new ArrayList<>();
+        List<AbstractPageBlock> chunkWords = new ArrayList<>();
+        List<ClassificationSection> chunkBlockList = new ArrayList<>();
+        List<ClassificationHeader> headers = new ArrayList<>();
+        List<ClassificationFooter> footers = new ArrayList<>();
        List<UnclassifiedText> unclassifiedTexts = new ArrayList<>();

-        AbstractTextContainer prev = null;
+        AbstractPageBlock prev = null;

        String lastHeadline = "";
-        Table previousTable = null;
-        for (Page page : document.getPages()) {
-            List<TextBlock> header = new ArrayList<>();
-            List<TextBlock> footer = new ArrayList<>();
-            List<TextBlock> unclassifiedText = new ArrayList<>();
-            for (AbstractTextContainer current : page.getTextBlocks()) {
+        TablePageBlock previousTable = null;
+        for (ClassificationPage page : document.getPages()) {
+            List<ClassificationTextBlock> header = new ArrayList<>();
+            List<ClassificationTextBlock> footer = new ArrayList<>();
+            List<ClassificationTextBlock> unclassifiedText = new ArrayList<>();
+            for (AbstractPageBlock current : page.getTextBlocks()) {

                if (current.getClassification() == null) {
                    continue;
@@ -54,22 +53,22 @@ public class SectionsBuilderService {
                current.setPage(page.getPageNumber());

                if (current.getClassification().equals("Header")) {
-                    header.add((TextBlock) current);
+                    header.add((ClassificationTextBlock) current);
                    continue;
                }

                if (current.getClassification().equals("Footer")) {
-                    footer.add((TextBlock) current);
+                    footer.add((ClassificationTextBlock) current);
                    continue;
                }

                if (current.getClassification().equals("Other")) {
-                    unclassifiedText.add((TextBlock) current);
+                    unclassifiedText.add((ClassificationTextBlock) current);
                    continue;
                }

                if (prev != null && current.getClassification().startsWith("H ") && !prev.getClassification().startsWith("H ") || !document.isHeadlines()) {
-                    Paragraph chunkBlock = buildTextBlock(chunkWords, lastHeadline);
+                    ClassificationSection chunkBlock = buildTextBlock(chunkWords, lastHeadline);
                    chunkBlock.setHeadline(lastHeadline);
                    if (document.isHeadlines()) {
                        lastHeadline = current.getText();
@@ -80,8 +79,7 @@ public class SectionsBuilderService {
                        previousTable = chunkBlock.getTables().get(chunkBlock.getTables().size() - 1);
                    }
                }
-                if (current instanceof Table) {
-                    Table table = (Table) current;
+                if (current instanceof TablePageBlock table) {
                    // Distribute header information for subsequent tables
                    mergeTableMetadata(table, previousTable);
                    previousTable = table;
@@ -91,69 +89,72 @@ public class SectionsBuilderService {
            }

            if (!header.isEmpty()) {
-                headers.add(new Header(header));
+                headers.add(new ClassificationHeader(header));
            }
            if (!footer.isEmpty()) {
-                footers.add(new Footer(footer));
+                footers.add(new ClassificationFooter(footer));
            }
            if (!unclassifiedText.isEmpty()) {
                unclassifiedTexts.add(new UnclassifiedText(unclassifiedText));
            }
        }

-        Paragraph chunkBlock = buildTextBlock(chunkWords, lastHeadline);
+        ClassificationSection chunkBlock = buildTextBlock(chunkWords, lastHeadline);
        chunkBlock.setHeadline(lastHeadline);
        chunkBlockList.add(chunkBlock);

-        document.setParagraphs(chunkBlockList);
+        document.setSections(chunkBlockList);
        document.setHeaders(headers);
        document.setFooters(footers);
        document.setUnclassifiedTexts(unclassifiedTexts);
    }


-    public void addImagesToSections(Document document) {
+    public void addImagesToSections(ClassificationDocument document) {

-        Map<Integer, List<Paragraph>> paragraphMap = new HashMap<>();
-        for (Paragraph paragraph : document.getParagraphs()) {
-            for (AbstractTextContainer container : paragraph.getPageBlocks()) {
-
-                paragraphMap.computeIfAbsent(container.getPage(), c -> new ArrayList<>()).add(paragraph);
+        Map<Integer, List<ClassificationSection>> sectionMap = new HashMap<>();
+        for (ClassificationSection section : document.getSections()) {
+            for (AbstractPageBlock container : section.getPageBlocks()) {

+                List<ClassificationSection> sectionsOnPage = sectionMap.computeIfAbsent(container.getPage(), c -> new ArrayList<>());
+                if (sectionsOnPage.contains(section)) {
+                    continue;
+                }
+                sectionsOnPage.add(section);
            }
        }

-        if (paragraphMap.isEmpty()) {
-            Paragraph paragraph = new Paragraph();
-            document.getParagraphs().add(paragraph);
-            paragraphMap.computeIfAbsent(1, x -> new ArrayList<>()).add(paragraph);
+        if (sectionMap.isEmpty()) {
+            ClassificationSection section = new ClassificationSection();
+            document.getSections().add(section);
+            sectionMap.computeIfAbsent(1, x -> new ArrayList<>()).add(section);
        }

        // first page is always a paragraph, else we can't process pages 1..N,
        // where N is the first found page with a paragraph
-        if (paragraphMap.get(1) == null) {
-            Paragraph paragraph = new Paragraph();
-            document.getParagraphs().add(paragraph);
-            paragraphMap.computeIfAbsent(1, x -> new ArrayList<>()).add(paragraph);
+        if (sectionMap.get(1) == null) {
+            ClassificationSection section = new ClassificationSection();
+            document.getSections().add(section);
+            sectionMap.computeIfAbsent(1, x -> new ArrayList<>()).add(section);
        }

-        for (Page page : document.getPages()) {
-            for (PdfImage image : page.getImages()) {
-                List<Paragraph> paragraphsOnPage = paragraphMap.get(page.getPageNumber());
-                if (paragraphsOnPage == null) {
+        for (ClassificationPage page : document.getPages()) {
+            for (ClassifiedImage image : page.getImages()) {
+                List<ClassificationSection> sectionsOnPage = sectionMap.get(page.getPageNumber());
+                if (sectionsOnPage == null) {
                    int i = page.getPageNumber();
-                    while (paragraphsOnPage == null) {
-                        paragraphsOnPage = paragraphMap.get(i);
+                    while (sectionsOnPage == null) {
+                        sectionsOnPage = sectionMap.get(i);
                        i--;
                    }
                }
-                for (Paragraph paragraph : paragraphsOnPage) {
+                for (ClassificationSection section : sectionsOnPage) {
                    Float xMin = null;
                    Float yMin = null;
                    Float xMax = null;
                    Float yMax = null;

-                    for (AbstractTextContainer abs : paragraph.getPageBlocks()) {
+                    for (AbstractPageBlock abs : section.getPageBlocks()) {
                        if (abs.getPage() != page.getPageNumber()) {
                            continue;
                        }
@@ -197,21 +198,21 @@ public class SectionsBuilderService {

                    if (xMin != null && xMax != null && yMin != null && yMax != null && image.getPosition().getX() >= xMin && image.getPosition()
                            .getX() <= xMax && image.getPosition().getY() >= yMin && image.getPosition().getY() <= yMax) {
-                        paragraph.getImages().add(image);
-                        image.setAppendedToParagraph(true);
+                        section.getImages().add(image);
+                        image.setAppendedToSection(true);
                    }
                }
-                if (!image.isAppendedToParagraph()) {
+                if (!image.isAppendedToSection()) {
                    log.debug("Image uses first paragraph");
-                    paragraphsOnPage.get(0).getImages().add(image);
-                    image.setAppendedToParagraph(true);
+                    sectionsOnPage.get(0).getImages().add(image);
+                    image.setAppendedToSection(true);
                }
            }
        }
    }


-    private void mergeTableMetadata(Table currentTable, Table previousTable) {
+    private void mergeTableMetadata(TablePageBlock currentTable, TablePageBlock previousTable) {

        // Distribute header information for subsequent tables
        if (previousTable != null && hasInvalidHeaderInformation(currentTable) && hasValidHeaderInformation(previousTable)) {
@@ -239,86 +240,44 @@ public class SectionsBuilderService {
    }


-    private Paragraph buildTextBlock(List<AbstractTextContainer> wordBlockList, String lastHeadline) {
+    private ClassificationSection buildTextBlock(List<AbstractPageBlock> wordBlockList, String lastHeadline) {

-        Paragraph paragraph = new Paragraph();
-        TextBlock textBlock = null;
+        ClassificationSection section = new ClassificationSection();

-        int pageBefore = -1;
-        boolean splitByTable = false;
+        for (AbstractPageBlock container : wordBlockList) {
+            if (container instanceof TablePageBlock table) {

-        Iterator<AbstractTextContainer> itty = wordBlockList.iterator();
-        boolean alreadyAdded = false;
-        AbstractTextContainer previous = null;
-        while (itty.hasNext()) {
-            AbstractTextContainer container = itty.next();
-
-            if (container instanceof Table) {
-                Table table = (Table) container;
-                splitByTable = true;
-
-                if (previous != null && previous.getText().startsWith("Table ")) {
-                    table.setHeadline(previous.getText());
+                if (lastHeadline == null || lastHeadline.isEmpty()) {
+                    table.setHeadline("Text in table");
                } else {
-                    if (lastHeadline == null || lastHeadline.isEmpty()) {
-                        table.setHeadline("Text in table");
-                    } else {
-                        table.setHeadline("Table in: " + lastHeadline);
-                    }
+                    table.setHeadline("TablePageBlock in: " + lastHeadline);
                }

-                if (textBlock != null && !alreadyAdded) {
-                    paragraph.getPageBlocks().add(textBlock);
-                    alreadyAdded = true;
-                }
-                paragraph.getPageBlocks().add(table);
+                section.getPageBlocks().add(table);
                continue;
            }

-            TextBlock wordBlock = (TextBlock) container;
-
-            if (textBlock == null) {
-                textBlock = new TextBlock(wordBlock.getMinX(), wordBlock.getMaxX(), wordBlock.getMinY(), wordBlock.getMaxY(), wordBlock.getSequences(), wordBlock.getRotation());
-                textBlock.setPage(wordBlock.getPage());
-            } else if (splitByTable) {
-                textBlock = new TextBlock(wordBlock.getMinX(), wordBlock.getMaxX(), wordBlock.getMinY(), wordBlock.getMaxY(), wordBlock.getSequences(), wordBlock.getRotation());
-                textBlock.setPage(wordBlock.getPage());
-                alreadyAdded = false;
-            } else if (pageBefore != -1 && wordBlock.getPage() != pageBefore) {
-                textBlock.setPage(pageBefore);
-                paragraph.getPageBlocks().add(textBlock);
-                textBlock = new TextBlock(wordBlock.getMinX(), wordBlock.getMaxX(), wordBlock.getMinY(), wordBlock.getMaxY(), wordBlock.getSequences(), wordBlock.getRotation());
-                textBlock.setPage(wordBlock.getPage());
-            } else {
-                TextBlock spatialEntity = textBlock.union(wordBlock);
-                textBlock.resize(spatialEntity.getMinX(), spatialEntity.getMinY(), spatialEntity.getWidth(), spatialEntity.getHeight());
-            }
-            pageBefore = wordBlock.getPage();
-            splitByTable = false;
-            previous = container;
+            ClassificationTextBlock wordBlock = (ClassificationTextBlock) container;
+            section.getPageBlocks().add(wordBlock);
        }
-
-        if (textBlock != null && !alreadyAdded) {
-            paragraph.getPageBlocks().add(textBlock);
-        }
-        return paragraph;
+        return section;
    }


-    private boolean hasValidHeaderInformation(Table table) {
+    private boolean hasValidHeaderInformation(TablePageBlock table) {

        return !hasInvalidHeaderInformation(table);
    }


-    private boolean hasInvalidHeaderInformation(Table table) {
+    private boolean hasInvalidHeaderInformation(TablePageBlock table) {

        return table.getRows().stream().flatMap(row -> row.stream().filter(cell -> CollectionUtils.isNotEmpty(cell.getHeaderCells()))).findAny().isEmpty();

    }


-    private List<Cell> getRowWithNonHeaderCells(Table table) {
+    private List<Cell> getRowWithNonHeaderCells(TablePageBlock table) {

        for (int i = table.getRowCount() - 1; i >= 0; i--) { // Non header rows are most likely at bottom of table
            List<Cell> row = table.getRows().get(i);
@@ -1,4 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.tableextraction.service;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.service;

 import java.awt.geom.Point2D;
 import java.util.ArrayList;
@@ -13,15 +13,15 @@ import java.util.stream.Collectors;

 import org.springframework.stereotype.Service;

-import com.iqser.red.service.redaction.v1.server.classification.model.Page;
-import com.iqser.red.service.redaction.v1.server.classification.model.TextBlock;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.AbstractTextContainer;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.Cell;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.CleanRulings;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.Rectangle;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.Ruling;
-import com.iqser.red.service.redaction.v1.server.tableextraction.model.Table;
-import com.iqser.red.service.redaction.v1.server.tableextraction.utils.Utils;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.AbstractPageBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.ClassificationPage;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table.Cell;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table.CleanRulings;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table.Rectangle;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table.Ruling;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table.TablePageBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.ClassificationTextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.utils.Utils;

@Service
 public class TableExtractionService {
@@ -73,20 +73,20 @@ public class TableExtractionService {
     * 90 -> UpperLeft
     * 180 -> UpperRight
     * 270 -> LowerRight
-     *
+     * <p>
     * DirAdj (Text direction adjusted) values can not be used here.
     *
     * @param cleanRulings The lines used to build the table.
-     * @param page Page object that contains textblocks and statistics.
+     * @param page         Page object that contains textblocks and statistics.
     */
-    public void extractTables(CleanRulings cleanRulings, Page page) {
+    public void extractTables(CleanRulings cleanRulings, ClassificationPage page) {

        List<Cell> cells = findCells(cleanRulings.getHorizontal(), cleanRulings.getVertical());

-        List<TextBlock> toBeRemoved = new ArrayList<>();
+        List<ClassificationTextBlock> toBeRemoved = new ArrayList<>();

-        for (AbstractTextContainer abstractTextContainer : page.getTextBlocks()) {
-            TextBlock textBlock = (TextBlock) abstractTextContainer;
+        for (AbstractPageBlock abstractPageBlock : page.getTextBlocks()) {
+            ClassificationTextBlock textBlock = (ClassificationTextBlock) abstractPageBlock;
            for (Cell cell : cells) {
                if (cell.hasMinimumSize() && cell.intersects(textBlock.getPdfMinX(),
                        textBlock.getPdfMinY(),
@@ -104,7 +104,7 @@ public class TableExtractionService {

        List<Rectangle> spreadsheetAreas = findSpreadsheetsFromCells(cells).stream().filter(r -> r.getWidth() > 0f && r.getHeight() > 0f).collect(Collectors.toList());

-        List<Table> tables = new ArrayList<>();
+        List<TablePageBlock> tables = new ArrayList<>();
        for (Rectangle area : spreadsheetAreas) {

            List<Cell> overlappingCells = new ArrayList<>();
@@ -113,16 +113,16 @@ public class TableExtractionService {
                    overlappingCells.add(c);
                }
            }
-            tables.add(new Table(overlappingCells, area, page.getRotation()));
+            tables.add(new TablePageBlock(overlappingCells, area, page.getRotation()));
        }

-        for (Table table : tables) {
+        for (TablePageBlock table : tables) {
            int position = -1;

-            Iterator<AbstractTextContainer> itty = page.getTextBlocks().iterator();
+            Iterator<AbstractPageBlock> itty = page.getTextBlocks().iterator();
            while (itty.hasNext()) {
-                AbstractTextContainer textBlock = itty.next();
-                if (textBlock instanceof TextBlock ? table.containsBlock((TextBlock) textBlock) : table.contains(textBlock) && position == -1) {
+                AbstractPageBlock textBlock = itty.next();
+                if (textBlock instanceof ClassificationTextBlock ? table.containsBlock((ClassificationTextBlock) textBlock) : table.contains(textBlock) && position == -1) {
                    position = page.getTextBlocks().indexOf(textBlock);
                }
            }
@@ -9,7 +9,7 @@
 * This program is free software under the LGPL (>=v2.1)
 * Read the file LICENSE.txt coming with the sources for details.
 */
-package com.iqser.red.service.redaction.v1.server.tableextraction.utils;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.utils;

 import java.awt.geom.Line2D;
 import java.awt.geom.Rectangle2D;
@@ -1,4 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.tableextraction.utils;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.utils;

 import java.io.File;
 import java.io.IOException;
@@ -23,9 +23,11 @@ public class FileUtils {
    /**
     * Deletes a file; logs a message with the reason if the deletion fails.
     * This method is null-safe.
+     *
     * @param file The file to delete.  Can be null.
     */
    public void deleteFile(File file) {
+
        if (file != null) {
            try {
                Files.deleteIfExists(file.toPath());
@@ -1,7 +1,7 @@
-package com.iqser.red.service.redaction.v1.server.classification.utils;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.utils;

 import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.Rectangle;
-import com.iqser.red.service.redaction.v1.server.classification.model.TextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.ClassificationTextBlock;

 import lombok.experimental.UtilityClass;

@@ -11,7 +11,7 @@ public final class PositionUtils {

    // TODO This currently uses pdf coord system. In the futher this should use java coord system.
    // Note: DirAdj (TextDirection Adjusted) can not be user for this.
-    public boolean isWithinBodyTextFrame(Rectangle btf, TextBlock textBlock) {
+    public boolean isWithinBodyTextFrame(Rectangle btf, ClassificationTextBlock textBlock) {

        if (btf == null || textBlock == null) {
            return false;
@@ -32,7 +32,7 @@ public final class PositionUtils {

    // TODO This currently uses pdf coord system. In the futher this should use java coord system.
    // Note: DirAdj (TextDirection Adjusted) can not be user for this.
-    public boolean isOverBodyTextFrame(Rectangle btf, TextBlock textBlock, int rotation) {
+    public boolean isOverBodyTextFrame(Rectangle btf, ClassificationTextBlock textBlock, int rotation) {

        if (btf == null || textBlock == null) {
            return false;
@@ -58,9 +58,10 @@ public final class PositionUtils {

    }

+
    // TODO This currently uses pdf coord system. In the futher this should use java coord system.
    // Note: DirAdj (TextDirection Adjusted) can not be user for this.
-    public boolean isUnderBodyTextFrame(Rectangle btf, TextBlock textBlock, int rotation) {
+    public boolean isUnderBodyTextFrame(Rectangle btf, ClassificationTextBlock textBlock, int rotation) {

        if (btf == null || textBlock == null) {
            return false;
@@ -86,9 +87,10 @@ public final class PositionUtils {

    }

+
    // TODO This currently uses pdf coord system. In the futher this should use java coord system.
    // Note: DirAdj (TextDirection Adjusted) can not be user for this.
-    public boolean isTouchingUnderBodyTextFrame(Rectangle btf, TextBlock textBlock) {
+    public boolean isTouchingUnderBodyTextFrame(Rectangle btf, ClassificationTextBlock textBlock) {

        //TODO Currently this is not working for rotated pages.

@@ -105,13 +107,13 @@ public final class PositionUtils {
    }


-    public float getHeightDifferenceBetweenChunkWordAndDocumentWord(TextBlock textBlock, Float documentMostPopularWordHeight) {
+    public float getHeightDifferenceBetweenChunkWordAndDocumentWord(ClassificationTextBlock textBlock, Float documentMostPopularWordHeight) {

        return textBlock.getMostPopularWordHeight() - documentMostPopularWordHeight;
    }


-    public Float getApproxLineCount(TextBlock textBlock) {
+    public Float getApproxLineCount(ClassificationTextBlock textBlock) {

        return textBlock.getHeight() / textBlock.getMostPopularWordHeight();
    }
@@ -1,4 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.tableextraction.utils;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.utils;

 import java.util.ArrayDeque;
 import java.util.Comparator;
@@ -1,9 +1,9 @@
-package com.iqser.red.service.redaction.v1.server.classification.utils;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.utils;

 import java.awt.geom.Line2D;
 import java.awt.geom.Point2D;

-import com.iqser.red.service.redaction.v1.server.tableextraction.model.Ruling;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table.Ruling;

 import lombok.experimental.UtilityClass;

@@ -13,7 +13,7 @@ public final class RulingTextDirAdjustUtil {
    /**
     * Converts a ruling (line of a table) the same way TextPositions are converted in PDFBox.
     * This will get the y position of the text, adjusted so that 0,0 is upper left and it is adjusted based on the text direction.
-     *
+     * <p>
     * See org.apache.pdfbox.text.TextPosition
     */
    public Line2D.Float convertToDirAdj(Ruling ruling, float dir, float pageWidth, float pageHeight) {
@@ -1,4 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.tableextraction.utils;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.classification.utils;

 import java.math.BigDecimal;
 import java.util.Comparator;
@@ -0,0 +1,19 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.data;
+
+import lombok.AccessLevel;
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.Data;
+import lombok.experimental.FieldDefaults;
+
+@Data
+@Builder
+@AllArgsConstructor
+@FieldDefaults(makeFinal = true, level = AccessLevel.PRIVATE)
+public class AtomicPositionBlockData {
+
+    Long id;
+    int[] stringIdxToPositionIdx;
+    float[][] positions;
+
+}
@@ -0,0 +1,23 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.data;
+
+import lombok.AccessLevel;
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.Data;
+import lombok.experimental.FieldDefaults;
+
+@Data
+@Builder
+@AllArgsConstructor
+@FieldDefaults(makeFinal = true, level = AccessLevel.PRIVATE)
+public class AtomicTextBlockData {
+
+    Long id;
+    Long page;
+    String searchText;
+    int numberOnPage;
+    int start;
+    int end;
+    int[] lineBreaks;
+
+}
@@ -0,0 +1,20 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.data;
+
+import lombok.AccessLevel;
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.Data;
+import lombok.experimental.FieldDefaults;
+
+@Data
+@Builder
+@AllArgsConstructor
+@FieldDefaults(makeFinal = true, level = AccessLevel.PRIVATE)
+public class DocumentData {
+
+    PageData[] pages;
+    AtomicTextBlockData[] atomicTextBlocks;
+    AtomicPositionBlockData[] atomicPositionBlocks;
+    DocumentTreeData documentTreeData;
+
+}
@@ -0,0 +1,93 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.data;
+
+import java.util.List;
+import java.util.Map;
+import java.util.stream.Stream;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.NodeType;
+
+import lombok.AccessLevel;
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.Data;
+import lombok.Getter;
+import lombok.NoArgsConstructor;
+import lombok.experimental.FieldDefaults;
+
+@Data
+@Builder
+@AllArgsConstructor
+@NoArgsConstructor
+@FieldDefaults(level = AccessLevel.PRIVATE)
+public class DocumentTreeData {
+
+    EntryData root;
+
+
+    public EntryData get(List<Integer> tocId) {
+
+        if (tocId.isEmpty()) {
+            return root;
+        }
+        EntryData entry = root.subEntries.get(tocId.get(0));
+        for (int id : tocId.subList(1, tocId.size())) {
+            entry = entry.subEntries.get(id);
+        }
+        return entry;
+    }
+
+
+    public Stream<EntryData> streamAllEntries() {
+
+        return Stream.concat(Stream.of(root), root.subEntries.stream()).flatMap(DocumentTreeData::flatten);
+    }
+
+
+    public String toString() {
+
+        return String.join("\n", streamAllEntries().map(EntryData::toString).toList());
+    }
+
+
+    private static Stream<EntryData> flatten(EntryData entry) {
+
+        return Stream.concat(Stream.of(entry), entry.subEntries.stream().flatMap(DocumentTreeData::flatten));
+    }
+
+
+    @Builder
+    @Getter
+    @AllArgsConstructor
+    @FieldDefaults(makeFinal = true, level = AccessLevel.PRIVATE)
+    public static class EntryData {
+
+        NodeType type;
+        int[] treeId;
+        Long[] atomicBlocks;
+        Long[] pages;
+        Map<String, String> properties;
+        List<EntryData> subEntries;
+
+
+        @Override
+        public String toString() {
+
+            StringBuilder sb = new StringBuilder();
+            sb.append("[");
+            for (int i : treeId) {
+                sb.append(i);
+                sb.append(",");
+            }
+            sb.delete(sb.length() - 1, sb.length());
+            sb.append("]: ");
+
+            sb.append(type);
+            sb.append(" atbs = ");
+            sb.append(atomicBlocks.length);
+
+            return sb.toString();
+        }
+
+    }
+
+}
@@ -0,0 +1,20 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.data;
+
+import lombok.AccessLevel;
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.Data;
+import lombok.experimental.FieldDefaults;
+
+@Data
+@Builder
+@AllArgsConstructor
+@FieldDefaults(makeFinal = true, level = AccessLevel.PRIVATE)
+public class PageData {
+
+    int number;
+    int height;
+    int width;
+    int rotation;
+
+}
@@ -0,0 +1,148 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.data.mapper;
+
+import java.awt.geom.Rectangle2D;
+import java.util.HashMap;
+import java.util.List;
+import java.util.Map;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.data.AtomicPositionBlockData;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.data.AtomicTextBlockData;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.data.DocumentData;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.data.DocumentTreeData;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.data.PageData;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.DocumentTree;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Document;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Image;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Page;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Section;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Table;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.TableCell;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.AtomicTextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.TextBlock;
+
+import lombok.experimental.UtilityClass;
+
+@UtilityClass
+public class DocumentDataMapper {
+
+    public DocumentData toDocumentData(Document document) {
+
+        List<AtomicTextBlockData> atomicTextBlockData = document.streamTerminalTextBlocksInOrder()
+                .flatMap(textBlock -> textBlock.getAtomicTextBlocks().stream())
+                .distinct()
+                .map(DocumentDataMapper::toAtomicTextBlockData)
+                .toList();
+
+        List<AtomicPositionBlockData> atomicPositionBlockData = document.streamTerminalTextBlocksInOrder()
+                .flatMap(textBlock -> textBlock.getAtomicTextBlocks().stream())
+                .distinct()
+                .map(DocumentDataMapper::toAtomicPositionBlockData)
+                .toList();
+
+        List<PageData> pageData = document.getPages().stream().map(DocumentDataMapper::toPageData).toList();
+        DocumentTreeData documentTreeData = toTableOfContentsData(document.getDocumentTree());
+        return DocumentData.builder()
+                .atomicTextBlocks(atomicTextBlockData.toArray(new AtomicTextBlockData[0]))
+                .atomicPositionBlocks(atomicPositionBlockData.toArray(new AtomicPositionBlockData[0]))
+                .pages(pageData.toArray(new PageData[0]))
+                .documentTreeData(documentTreeData)
+                .build();
+    }
+
+
+    private DocumentTreeData toTableOfContentsData(DocumentTree documentTree) {
+
+        return new DocumentTreeData(toEntryData(documentTree.getRoot()));
+    }
+
+
+    private DocumentTreeData.EntryData toEntryData(DocumentTree.Entry entry) {
+
+        Long[] atomicTextBlocks;
+
+        if (entry.getNode().isLeaf()) {
+            atomicTextBlocks = toAtomicTextBlockIds(entry.getNode().getLeafTextBlock());
+        } else {
+            atomicTextBlocks = new Long[]{};
+        }
+
+        Map<String, String> properties = switch (entry.getType()) {
+            case TABLE -> PropertiesMapper.buildTableProperties((Table) entry.getNode());
+            case TABLE_CELL -> PropertiesMapper.buildTableCellProperties((TableCell) entry.getNode());
+            case IMAGE -> PropertiesMapper.buildImageProperties((Image) entry.getNode());
+            case SECTION -> PropertiesMapper.buildSectionProperties((Section) entry.getNode());
+            default -> new HashMap<>();
+        };
+
+        return DocumentTreeData.EntryData.builder()
+                .treeId(toPrimitiveIntArray(entry.getTocId()))
+                .subEntries(entry.getChildren().stream().map(DocumentDataMapper::toEntryData).toList())
+                .type(entry.getType())
+                .atomicBlocks(atomicTextBlocks)
+                .pages(entry.getNode().getPages().stream().map(Page::getNumber).map(Integer::longValue).toArray(Long[]::new))
+                .properties(properties)
+                .build();
+    }
+
+
+    private Long[] toAtomicTextBlockIds(TextBlock textBlock) {
+
+        return textBlock.getAtomicTextBlocks().stream().map(AtomicTextBlock::getId).toArray(Long[]::new);
+    }
+
+
+    private PageData toPageData(Page p) {
+
+        return PageData.builder().rotation(p.getRotation()).height(p.getHeight()).width(p.getWidth()).number(p.getNumber()).build();
+    }
+
+
+    private AtomicTextBlockData toAtomicTextBlockData(AtomicTextBlock atomicTextBlock) {
+
+        return AtomicTextBlockData.builder()
+                .id(atomicTextBlock.getId())
+                .page(atomicTextBlock.getPage().getNumber().longValue())
+                .searchText(atomicTextBlock.getSearchText())
+                .numberOnPage(atomicTextBlock.getNumberOnPage())
+                .start(atomicTextBlock.getBoundary().start())
+                .end(atomicTextBlock.getBoundary().end())
+                .lineBreaks(toPrimitiveIntArray(atomicTextBlock.getLineBreaks()))
+                .build();
+    }
+
+
+    private AtomicPositionBlockData toAtomicPositionBlockData(AtomicTextBlock atomicTextBlock) {
+
+        return AtomicPositionBlockData.builder()
+                .id(atomicTextBlock.getId())
+                .positions(toPrimitiveFloatMatrix(atomicTextBlock.getPositions()))
+                .stringIdxToPositionIdx(toPrimitiveIntArray(atomicTextBlock.getStringIdxToPositionIdx()))
+                .build();
+    }
+
+
+    private float[][] toPrimitiveFloatMatrix(List<Rectangle2D> positions) {
+
+        float[][] positionMatrix = new float[positions.size()][];
+        for (int i = 0; i < positions.size(); i++) {
+            float[] singlePositions = new float[4];
+            singlePositions[0] = (float) positions.get(i).getMinX();
+            singlePositions[1] = (float) positions.get(i).getMinY();
+            singlePositions[2] = (float) positions.get(i).getWidth();
+            singlePositions[3] = (float) positions.get(i).getHeight();
+            positionMatrix[i] = singlePositions;
+        }
+        return positionMatrix;
+    }
+
+
+    private int[] toPrimitiveIntArray(List<Integer> list) {
+
+        int[] array = new int[list.size()];
+        for (int i = 0; i < list.size(); i++) {
+            array[i] = list.get(i);
+        }
+        return array;
+    }
+
+}
@@ -0,0 +1,227 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.data.mapper;
+
+import static com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.NodeType.FOOTER;
+import static com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.NodeType.HEADER;
+
+import java.awt.geom.Rectangle2D;
+import java.util.Arrays;
+import java.util.HashSet;
+import java.util.LinkedList;
+import java.util.List;
+import java.util.Map;
+import java.util.NoSuchElementException;
+
+import com.google.common.primitives.Ints;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.data.AtomicPositionBlockData;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.data.AtomicTextBlockData;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.data.DocumentData;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.data.DocumentTreeData;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.data.PageData;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.Boundary;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.DocumentTree;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Document;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Footer;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Header;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Headline;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Image;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Page;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Paragraph;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Section;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.SemanticNode;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Table;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.TableCell;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.AtomicTextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.TextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.TextBlockCollector;
+
+import lombok.experimental.UtilityClass;
+
+@UtilityClass
+public class DocumentGraphMapper {
+
+    public Document toDocumentGraph(DocumentData documentData) {
+
+        Document document = new Document();
+        Context context = new Context(documentData,
+                new DocumentTree(document),
+                new LinkedList<>(),
+                new LinkedList<>(),
+                Arrays.stream(documentData.getAtomicTextBlocks()).toList(),
+                Arrays.stream(documentData.getAtomicPositionBlocks()).toList());
+
+        context.pages.addAll(Arrays.stream(documentData.getPages()).map(DocumentGraphMapper::buildPage).toList());
+
+        context.documentTree.getRoot().getChildren().addAll(buildEntries(documentData.getDocumentTreeData().getRoot().getSubEntries(), context));
+
+        document.setDocumentTree(context.documentTree);
+        document.setPages(new HashSet<>(context.pages));
+        document.setNumberOfPages(documentData.getPages().length);
+
+        document.setTextBlock(document.buildTextBlock());
+        return document;
+    }
+
+
+    private List<DocumentTree.Entry> buildEntries(List<DocumentTreeData.EntryData> entries, Context context) {
+
+        List<DocumentTree.Entry> newEntries = new LinkedList<>();
+        for (DocumentTreeData.EntryData entryData : entries) {
+
+            boolean terminal = isTerminal(entryData);
+            List<Page> pages = Arrays.stream(entryData.getPages()).map(pageNumber -> getPage(pageNumber, context)).toList();
+
+            SemanticNode node = switch (entryData.getType()) {
+                case SECTION -> buildSection(context, entryData.getProperties());
+                case PARAGRAPH -> buildParagraph(context, terminal);
+                case HEADLINE -> buildHeadline(context, terminal);
+                case HEADER -> buildHeader(context, terminal);
+                case FOOTER -> buildFooter(context, terminal);
+                case TABLE -> buildTable(context, entryData.getProperties());
+                case TABLE_CELL -> buildTableCell(context, entryData.getProperties(), terminal);
+                case IMAGE -> buildImage(context, entryData.getProperties(), entryData.getPages());
+                default -> throw new UnsupportedOperationException("Not yet implemented for type " + entryData.getType());
+            };
+
+            if (node.isLeaf()) {
+                TextBlock textBlock = toTextBlock(entryData.getAtomicBlocks(), context, node);
+                node.setLeafTextBlock(textBlock);
+            }
+            List<Integer> tocId = Arrays.stream(entryData.getTreeId()).boxed().toList();
+            node.setTreeId(tocId);
+
+            if (entryData.getType() == HEADER) {
+                pages.forEach(page -> page.setHeader((Header) node));
+            } else if (entryData.getType() == FOOTER) {
+                pages.forEach(page -> page.setFooter((Footer) node));
+            } else {
+                pages.forEach(page -> page.getMainBody().add(node));
+            }
+            newEntries.add(DocumentTree.Entry.builder().tocId(tocId).type(entryData.getType()).children(buildEntries(entryData.getSubEntries(), context)).node(node).build());
+        }
+        return newEntries;
+    }
+
+
+    private Headline buildHeadline(Context context, boolean terminal) {
+
+        return Headline.builder().leaf(terminal).documentTree(context.documentTree()).build();
+    }
+
+
+    private static boolean isTerminal(DocumentTreeData.EntryData entryData) {
+
+        return entryData.getAtomicBlocks().length > 0;
+    }
+
+
+    private Image buildImage(Context context, Map<String, String> properties, Long[] pageNumbers) {
+
+        assert pageNumbers.length == 1;
+        Page page = getPage(pageNumbers[0], context);
+        var builder = Image.builder();
+        PropertiesMapper.parseImageProperties(properties, builder);
+        return builder.documentTree(context.documentTree()).page(page).build();
+    }
+
+
+    private TableCell buildTableCell(Context context, Map<String, String> properties, boolean terminal) {
+
+        TableCell.TableCellBuilder builder = TableCell.builder();
+        PropertiesMapper.parseTableCellProperties(properties, builder);
+        return builder.leaf(terminal).documentTree(context.documentTree()).build();
+    }
+
+
+    private Table buildTable(Context context, Map<String, String> properties) {
+
+        Table.TableBuilder builder = Table.builder();
+        PropertiesMapper.parseTableProperties(properties, builder);
+        return builder.documentTree(context.documentTree()).build();
+    }
+
+
+    private Footer buildFooter(Context context, boolean terminal) {
+
+        return Footer.builder().leaf(terminal).documentTree(context.documentTree()).build();
+    }
+
+
+    private Header buildHeader(Context context, boolean terminal) {
+
+        return Header.builder().leaf(terminal).documentTree(context.documentTree()).build();
+    }
+
+
+    private Section buildSection(Context context, Map<String, String> properties) {
+
+        var builder = Section.builder();
+        PropertiesMapper.parseSectionProperties(properties, builder);
+        return builder.documentTree(context.documentTree()).build();
+    }
+
+
+    private Paragraph buildParagraph(Context context, boolean terminal) {
+
+        return Paragraph.builder().leaf(terminal).documentTree(context.documentTree()).build();
+    }
+
+
+    private TextBlock toTextBlock(Long[] atomicTextBlockIds, Context context, SemanticNode parent) {
+
+        return Arrays.stream(atomicTextBlockIds)
+                .map(atomicTextBlockId -> toAtomicTextBlock(context.atomicTextBlockData.get(Math.toIntExact(atomicTextBlockId)),
+                        context.atomicPositionBlockData.get(Math.toIntExact(atomicTextBlockId)),
+                        parent,
+                        context))
+                .collect(new TextBlockCollector());
+    }
+
+
+    private Page buildPage(PageData p) {
+
+        return Page.builder().rotation(p.getRotation()).height(p.getHeight()).width(p.getWidth()).number(p.getNumber()).mainBody(new LinkedList<>()).build();
+    }
+
+
+    private AtomicTextBlock toAtomicTextBlock(AtomicTextBlockData atomicTextBlockData, AtomicPositionBlockData atomicPositionBlockData, SemanticNode parent, Context context) {
+
+        return AtomicTextBlock.builder()
+                .id(atomicTextBlockData.getId())
+                .numberOnPage(atomicTextBlockData.getNumberOnPage())
+                .page(getPage(atomicTextBlockData.getPage(), context))
+                .boundary(new Boundary(atomicTextBlockData.getStart(), atomicTextBlockData.getEnd()))
+                .searchText(atomicTextBlockData.getSearchText())
+                .lineBreaks(Ints.asList(atomicTextBlockData.getLineBreaks()))
+                .stringIdxToPositionIdx(Ints.asList(atomicPositionBlockData.getStringIdxToPositionIdx()))
+                .positions(toRectangle2DList(atomicPositionBlockData.getPositions()))
+                .parent(parent)
+                .build();
+    }
+
+
+    private static List<Rectangle2D> toRectangle2DList(float[][] positions) {
+
+        return Arrays.stream(positions).map(floatArr -> (Rectangle2D) new Rectangle2D.Float(floatArr[0], floatArr[1], floatArr[2], floatArr[3])).toList();
+    }
+
+
+    private Page getPage(Long pageIndex, Context context) {
+
+        return context.pages.stream()
+                .filter(page -> page.getNumber() == Math.toIntExact(pageIndex))
+                .findFirst()
+                .orElseThrow(() -> new NoSuchElementException(String.format("ClassificationPage with number %d not found", pageIndex)));
+    }
+
+
+    record Context(
+            DocumentData layoutParsingModel,
+            DocumentTree documentTree,
+            List<Page> pages,
+            List<Section> sections,
+            List<AtomicTextBlockData> atomicTextBlockData,
+            List<AtomicPositionBlockData> atomicPositionBlockData) {
+
+    }
+
+}
@@ -0,0 +1,104 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.data.mapper;
+
+import static com.iqser.red.service.redaction.v1.server.layoutparsing.document.utils.RectangleMapper.parseRectangle2D;
+
+import java.util.HashMap;
+import java.util.Map;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Image;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.ImageType;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Section;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Table;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.TableCell;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.utils.RectangleMapper;
+
+public class PropertiesMapper {
+
+    public static Map<String, String> buildImageProperties(Image image) {
+
+        Map<String, String> properties = new HashMap<>();
+        properties.put("imageType", image.getImageType().toString());
+        properties.put("transparency", String.valueOf(image.isTransparent()));
+        properties.put("position", RectangleMapper.toString(image.getPosition()));
+        properties.put("id", image.getId());
+        return properties;
+    }
+
+
+    public static Map<String, String> buildTableCellProperties(TableCell tableCell) {
+
+        Map<String, String> properties = new HashMap<>();
+        properties.put("row", String.valueOf(tableCell.getRow()));
+        properties.put("col", String.valueOf(tableCell.getCol()));
+        properties.put("header", String.valueOf(tableCell.isHeader()));
+
+        if (tableCell.getPages().size() > 1 || tableCell.getBBox().keySet().size() > 1) {
+            throw new IllegalArgumentException("TableCell can only occur on a single page!");
+        }
+        String bBoxString = RectangleMapper.toString(tableCell.getBBox().get(tableCell.getPages().stream().findFirst().get()));
+        properties.put("bBox", bBoxString);
+
+        return properties;
+    }
+
+
+    public static Map<String, String> buildTableProperties(Table table) {
+
+        Map<String, String> properties = new HashMap<>();
+        properties.put("numberOfRows", String.valueOf(table.getNumberOfRows()));
+        properties.put("numberOfCols", String.valueOf(table.getNumberOfCols()));
+        return properties;
+    }
+
+
+    public static Map<String, String> buildSectionProperties(Section node) {
+
+        Map<String, String> properties = new HashMap<>();
+        properties.put("excludesTables", String.valueOf(node.isExcludesTables()));
+        return properties;
+    }
+
+
+    public static void parseImageProperties(Map<String, String> properties, Image.ImageBuilder builder) {
+
+        builder.imageType(parseImageType(properties.get("imageType")));
+        builder.transparent(Boolean.parseBoolean(properties.get("transparency")));
+        builder.position(parseRectangle2D(properties.get("position")));
+        builder.id(properties.get("id"));
+    }
+
+
+    public static void parseTableCellProperties(Map<String, String> properties, TableCell.TableCellBuilder builder) {
+
+        builder.row(Integer.parseInt(properties.get("row")));
+        builder.col(Integer.parseInt(properties.get("col")));
+        builder.header(Boolean.parseBoolean(properties.get("header")));
+        builder.bBox(parseRectangle2D(properties.get("bBox")));
+    }
+
+
+    public static void parseTableProperties(Map<String, String> properties, Table.TableBuilder builder) {
+
+        builder.numberOfRows(Integer.parseInt(properties.get("numberOfRows")));
+        builder.numberOfCols(Integer.parseInt(properties.get("numberOfCols")));
+    }
+
+
+    public static void parseSectionProperties(Map<String, String> properties, Section.SectionBuilder builder) {
+
+        builder.excludesTables(Boolean.parseBoolean(properties.get("excludesTables")));
+    }
+
+
+    public static ImageType parseImageType(String imageType) {
+
+        return switch (imageType.toLowerCase()) {
+            case "logo" -> ImageType.LOGO;
+            case "formula" -> ImageType.FORMULA;
+            case "signature" -> ImageType.SIGNATURE;
+            case "ocr" -> ImageType.OCR;
+            default -> ImageType.OTHER;
+        };
+    }
+
+}
@@ -0,0 +1,249 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.factory;
+
+import static java.lang.String.format;
+import static java.util.stream.Collectors.groupingBy;
+import static java.util.stream.Collectors.toList;
+
+import java.awt.geom.Rectangle2D;
+import java.util.HashMap;
+import java.util.LinkedList;
+import java.util.List;
+import java.util.Map;
+import java.util.NoSuchElementException;
+import java.util.Set;
+import java.util.concurrent.atomic.AtomicInteger;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.AbstractPageBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.ClassificationDocument;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.ClassificationFooter;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.ClassificationHeader;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.ClassificationPage;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.image.ClassifiedImage;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.ClassificationTextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.DocumentTree;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Document;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Footer;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Header;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Headline;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Image;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.NodeType;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Page;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Paragraph;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Section;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.SemanticNode;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.AtomicTextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.utils.TextPositionOperations;
+import com.iqser.red.service.redaction.v1.server.redaction.utils.IdBuilder;
+
+import lombok.AccessLevel;
+import lombok.Builder;
+import lombok.Getter;
+import lombok.experimental.FieldDefaults;
+import lombok.experimental.UtilityClass;
+
+@UtilityClass
+public class DocumentGraphFactory {
+
+    public Document buildDocumentGraph(ClassificationDocument document) {
+
+        TextBlockFactory textBlockFactory = new TextBlockFactory();
+        Document documentGraph = new Document();
+        Context context = new Context(new DocumentTree(documentGraph), new HashMap<>(), new LinkedList<>(), new LinkedList<>(), textBlockFactory);
+
+        document.getPages().stream().map(DocumentGraphFactory::buildPage).forEach(page -> context.getPages().put(page, new AtomicInteger(1)));
+        document.getSections().stream().flatMap(section -> section.getImages().stream()).forEach(image -> context.getImages().add(image));
+        addSections(document, context);
+        addHeaderAndFooterToEachPage(document, context);
+
+        documentGraph.setNumberOfPages(context.pages.size());
+        documentGraph.setPages(context.pages.keySet());
+        documentGraph.setDocumentTree(context.documentTree);
+        documentGraph.setTextBlock(documentGraph.buildTextBlock());
+        return documentGraph;
+    }
+
+
+    private void addSections(ClassificationDocument document, Context context) {
+
+        document.getSections().forEach(section -> SectionNodeFactory.addSection(null, section.getPageBlocks(), section.getImages(), context));
+    }
+
+
+    public static boolean pageBlockIsHeadline(AbstractPageBlock abstractPageBlock) {
+
+        return abstractPageBlock instanceof ClassificationTextBlock && abstractPageBlock.getClassification() != null && abstractPageBlock.getClassification().startsWith("H");
+    }
+
+
+    public void addParagraphOrHeadline(SemanticNode parentNode, ClassificationTextBlock originalTextBlock, Context context, List<ClassificationTextBlock> textBlocksToMerge) {
+
+        Page page = context.getPage(originalTextBlock.getPage());
+
+        SemanticNode node;
+        if (pageBlockIsHeadline(originalTextBlock)) {
+            node = Headline.builder().documentTree(context.getDocumentTree()).build();
+        } else {
+            node = Paragraph.builder().documentTree(context.getDocumentTree()).build();
+        }
+
+        page.getMainBody().add(node);
+
+        List<ClassificationTextBlock> textBlocks = new LinkedList<>(textBlocksToMerge);
+        textBlocks.add(originalTextBlock);
+        AtomicTextBlock textBlock = context.textBlockFactory.buildAtomicTextBlock(TextPositionOperations.mergeAndSortTextPositionSequenceByYThenX(textBlocks), node, context, page);
+
+        if (node instanceof Headline headline) {
+            List<Integer> tocId = context.documentTree.createNewChildEntryAndReturnId(parentNode.getTreeId(), NodeType.HEADLINE, node);
+            headline.setLeafTextBlock(textBlock);
+            headline.setTreeId(tocId);
+        }
+        if (node instanceof Paragraph paragraph) {
+            List<Integer> tocId = context.documentTree.createNewChildEntryAndReturnId(parentNode.getTreeId(), NodeType.PARAGRAPH, node);
+            paragraph.setLeafTextBlock(textBlock);
+            paragraph.setTreeId(tocId);
+        }
+    }
+
+
+    public void addImage(Section section, ClassifiedImage image, Context context) {
+
+        Rectangle2D position = image.getPosition();
+        Page page = context.getPage(image.getPage());
+        Image imageNode = Image.builder()
+                .id(IdBuilder.buildId(Set.of(page), List.of(position)))
+                .imageType(image.getImageType())
+                .position(position)
+                .transparent(image.isHasTransparency())
+                .page(page)
+                .documentTree(context.getDocumentTree())
+                .build();
+        page.getMainBody().add(imageNode);
+
+        List<Integer> tocId = context.getDocumentTree().createNewChildEntryAndReturnId(section.getTreeId(), NodeType.IMAGE, imageNode);
+        imageNode.setTreeId(tocId);
+    }
+
+
+    private void addHeaderAndFooterToEachPage(ClassificationDocument document, Context context) {
+
+        Map<Integer, List<ClassificationTextBlock>> headers = document.getHeaders()
+                .stream()
+                .map(ClassificationHeader::getTextBlocks)
+                .flatMap(List::stream)
+                .collect(groupingBy(AbstractPageBlock::getPage, toList()));
+
+        Map<Integer, List<ClassificationTextBlock>> footers = document.getFooters()
+                .stream()
+                .map(ClassificationFooter::getTextBlocks)
+                .flatMap(List::stream)
+                .collect(groupingBy(AbstractPageBlock::getPage, toList()));
+
+        for (int pageIndex = 1; pageIndex <= document.getPages().size(); pageIndex++) {
+            if (headers.containsKey(pageIndex)) {
+                addHeader(headers.get(pageIndex), context);
+            } else {
+                addEmptyHeader(pageIndex, context);
+            }
+        }
+
+        for (int pageIndex = 1; pageIndex <= document.getPages().size(); pageIndex++) {
+            if (footers.containsKey(pageIndex)) {
+                addFooter(footers.get(pageIndex), context);
+            } else {
+                addEmptyFooter(pageIndex, context);
+            }
+        }
+    }
+
+
+    private void addFooter(List<ClassificationTextBlock> textBlocks, Context context) {
+
+        Page page = context.getPage(textBlocks.get(0).getPage());
+        Footer footer = Footer.builder().documentTree(context.getDocumentTree()).build();
+        AtomicTextBlock textBlock = context.textBlockFactory.buildAtomicTextBlock(TextPositionOperations.mergeAndSortTextPositionSequenceByYThenX(textBlocks),
+                footer,
+                context,
+                page);
+        List<Integer> tocId = context.getDocumentTree().createNewMainEntryAndReturnId(NodeType.FOOTER, footer);
+        footer.setTreeId(tocId);
+        footer.setLeafTextBlock(textBlock);
+        page.setFooter(footer);
+    }
+
+
+    public void addHeader(List<ClassificationTextBlock> textBlocks, Context context) {
+
+        Page page = context.getPage(textBlocks.get(0).getPage());
+        Header header = Header.builder().documentTree(context.getDocumentTree()).build();
+        AtomicTextBlock textBlock = context.textBlockFactory.buildAtomicTextBlock(TextPositionOperations.mergeAndSortTextPositionSequenceByYThenX(textBlocks),
+                header,
+                context,
+                0,
+                page);
+        List<Integer> tocId = context.getDocumentTree().createNewMainEntryAndReturnId(NodeType.HEADER, header);
+        header.setTreeId(tocId);
+        header.setLeafTextBlock(textBlock);
+        page.setHeader(header);
+    }
+
+
+    private void addEmptyFooter(int pageIndex, Context context) {
+
+        Page page = context.getPage(pageIndex);
+        Footer footer = Footer.builder().documentTree(context.getDocumentTree()).build();
+        AtomicTextBlock textBlock = context.textBlockFactory.emptyTextBlock(footer, context, page);
+        List<Integer> tocId = context.getDocumentTree().createNewMainEntryAndReturnId(NodeType.FOOTER, footer);
+        footer.setTreeId(tocId);
+        footer.setLeafTextBlock(textBlock);
+        page.setFooter(footer);
+    }
+
+
+    private void addEmptyHeader(int pageIndex, Context context) {
+
+        Page page = context.getPage(pageIndex);
+        Header header = Header.builder().documentTree(context.getDocumentTree()).build();
+        AtomicTextBlock textBlock = context.textBlockFactory.emptyTextBlock(header, 0, page);
+        List<Integer> tocId = context.getDocumentTree().createNewMainEntryAndReturnId(NodeType.HEADER, header);
+        header.setTreeId(tocId);
+        header.setLeafTextBlock(textBlock);
+        page.setHeader(header);
+    }
+
+
+    private Page buildPage(ClassificationPage p) {
+
+        return Page.builder()
+                .height((int) p.getPageHeight())
+                .width((int) p.getPageWidth())
+                .number(p.getPageNumber())
+                .rotation(p.getRotation())
+                .mainBody(new LinkedList<>())
+                .build();
+    }
+
+
+    @Getter
+    @Builder
+    @FieldDefaults(makeFinal = true, level = AccessLevel.PRIVATE)
+    static final class Context {
+
+        DocumentTree documentTree;
+        Map<Page, AtomicInteger> pages;
+        List<Section> sections;
+        List<ClassifiedImage> images;
+        TextBlockFactory textBlockFactory;
+
+
+        public Page getPage(int pageIndex) {
+
+            return pages.keySet()
+                    .stream()
+                    .filter(page -> page.getNumber() == pageIndex)
+                    .findFirst()
+                    .orElseThrow(() -> new NoSuchElementException(format("ClassificationPage with number %d not found", pageIndex)));
+        }
+
+    }
+
+}
@@ -0,0 +1,21 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.factory;
+
+import java.awt.geom.Rectangle2D;
+import java.util.List;
+
+import lombok.AccessLevel;
+import lombok.Builder;
+import lombok.Getter;
+import lombok.experimental.FieldDefaults;
+
+@Builder
+@Getter
+@FieldDefaults(makeFinal = true, level = AccessLevel.PRIVATE)
+public class SearchTextWithTextPositionDto {
+
+    String searchText;
+    List<Integer> lineBreaks;
+    List<Integer> stringCoordsToPositionCoords;
+    List<Rectangle2D> positions;
+
+}
@@ -0,0 +1,159 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.factory;
+
+import java.awt.geom.AffineTransform;
+import java.awt.geom.Rectangle2D;
+import java.util.Collections;
+import java.util.LinkedList;
+import java.util.List;
+import java.util.Objects;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.RedTextPosition;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.TextDirection;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.TextPositionSequence;
+
+public class SearchTextWithTextPositionFactory {
+
+    public static final int HEIGHT_PADDING = 2;
+
+
+    public static SearchTextWithTextPositionDto buildSearchTextToTextPositionModel(List<TextPositionSequence> sequences) {
+
+        if (sequences.isEmpty() || sequences.stream().allMatch(sequence -> sequence.getTextPositions().isEmpty())) {
+            return SearchTextWithTextPositionDto.builder()
+                    .searchText("")
+                    .lineBreaks(Collections.emptyList())
+                    .positions(Collections.emptyList())
+                    .stringCoordsToPositionCoords(Collections.emptyList())
+                    .build();
+        }
+
+        List<Integer> stringIdxToPositionIdx = new LinkedList<>();
+        List<Integer> lineBreaksStringIdx = new LinkedList<>();
+        StringBuilder sb = new StringBuilder();
+
+        int stringIdx = 0;
+        int positionIdx = 0;
+        int lastHyphenIdx = -3;
+
+        RedTextPosition currentTextPosition = sequences.get(0).getTextPositions().get(0);
+        RedTextPosition previousTextPosition = RedTextPosition.builder().unicode(" ").position(currentTextPosition.getPosition()).build();
+
+        for (TextPositionSequence word : sequences) {
+            for (int i = 0; i < word.getTextPositions().size(); ++i) {
+
+                currentTextPosition = word.getTextPositions().get(i);
+
+                if (isLineBreak(currentTextPosition, previousTextPosition)) {
+
+                    if (stringIdx - lastHyphenIdx < 3) {
+                        sb.delete(lastHyphenIdx, sb.length());
+                        stringIdxToPositionIdx = stringIdxToPositionIdx.subList(0, lastHyphenIdx);
+                        stringIdx = lastHyphenIdx;
+                        lastHyphenIdx = -3;
+                    }
+                    lineBreaksStringIdx.add(stringIdx);
+                }
+                if (!isRepeatedWhitespace(currentTextPosition.getUnicode(), previousTextPosition.getUnicode())) {
+
+                    if (isHyphen(currentTextPosition.getUnicode())) {
+                        lastHyphenIdx = stringIdx;
+                    }
+                    sb.append(currentTextPosition.getUnicode());
+
+                    // unicode characters with more than 16 bit encoding have a length > 1 in java strings
+                    for (int j = 0; j < currentTextPosition.getUnicode().length(); j++) {
+                        stringIdxToPositionIdx.add(positionIdx);
+                    }
+                    stringIdx += currentTextPosition.getUnicode().length();
+                }
+
+                previousTextPosition = currentTextPosition;
+                ++positionIdx;
+            }
+
+            previousTextPosition = RedTextPosition.builder().unicode(" ").position(previousTextPosition.getPosition()).build();
+            sb.append(" ");
+            stringIdxToPositionIdx.add(positionIdx);
+            ++stringIdx;
+        }
+
+        assert sb.length() == stringIdxToPositionIdx.size();
+
+        List<Rectangle2D> positions = sequences.stream()
+                .flatMap(sequence -> sequence.getTextPositions().stream().map(textPosition -> mapRedTextPositionToInitialUserSpace(textPosition, sequence)))
+                .toList();
+
+        return SearchTextWithTextPositionDto.builder()
+                .searchText(sb.toString())
+                .lineBreaks(lineBreaksStringIdx)
+                .stringCoordsToPositionCoords(stringIdxToPositionIdx)
+                .positions(positions)
+                .build();
+    }
+
+
+    private static boolean isLineBreak(RedTextPosition currentTextPosition, RedTextPosition previousTextPosition) {
+
+        return Objects.equals(currentTextPosition.getUnicode(), "\n") || isDeltaYLargerThanTextHeight(currentTextPosition, previousTextPosition);
+    }
+
+
+    private static boolean isDeltaYLargerThanTextHeight(RedTextPosition currentPosition, RedTextPosition previousPosition) {
+
+        if (previousPosition == null) {
+            return false;
+        }
+
+        float deltaY = Math.abs(currentPosition.getYDirAdj() - previousPosition.getYDirAdj());
+        return deltaY >= currentPosition.getHeightDir();
+    }
+
+
+    private static boolean isRepeatedWhitespace(String currentUnicode, String previousUnicode) {
+
+        return Objects.equals(previousUnicode, " ") && Objects.equals(currentUnicode, " ");
+    }
+
+
+    private static boolean isHyphen(String unicodeCharacter) {
+
+        return Objects.equals(unicodeCharacter, "-") || //
+                Objects.equals(unicodeCharacter, "~") || //
+                Objects.equals(unicodeCharacter, "‐") || //
+                Objects.equals(unicodeCharacter, "‒") || //
+                Objects.equals(unicodeCharacter, "⁻") || //
+                Objects.equals(unicodeCharacter, "−") || //
+                Objects.equals(unicodeCharacter, "﹣") || //
+                Objects.equals(unicodeCharacter, "゠") || //
+                Objects.equals(unicodeCharacter, "⁓") || //
+                Objects.equals(unicodeCharacter, "‑") || //
+                Objects.equals(unicodeCharacter, "\u00AD");
+    }
+
+
+    private static Rectangle2D mapRedTextPositionToInitialUserSpace(RedTextPosition textPosition, TextPositionSequence sequence) {
+
+        float textHeight = sequence.getTextHeight() + HEIGHT_PADDING;
+        Rectangle2D rectangle2D = new Rectangle2D.Double(textPosition.getXDirAdj(),
+                textPosition.getYDirAdj() - textHeight,
+                textPosition.getWidthDirAdj(),
+                textHeight + HEIGHT_PADDING);
+
+        AffineTransform transform = new AffineTransform();
+
+        if (sequence.getDir() == TextDirection.ZERO || sequence.getDir() == TextDirection.HALF_CIRCLE) {
+            transform.rotate(sequence.getDir().getRadians(), sequence.getPageWidth() / 2f, sequence.getPageHeight() / 2f);
+            transform.translate(0f, sequence.getPageHeight());
+        } else if (sequence.getDir() == TextDirection.QUARTER_CIRCLE) {
+            transform.rotate(sequence.getDir().getRadians(), sequence.getPageWidth() / 2f, sequence.getPageWidth() / 2f);
+            transform.translate(0f, sequence.getPageWidth());
+        } else {
+            transform.rotate(sequence.getDir().getRadians(), sequence.getPageHeight() / 2f, sequence.getPageHeight() / 2f);
+            transform.translate(0f, sequence.getPageWidth());
+        }
+        transform.scale(1., -1.);
+
+        return transform.createTransformedShape(rectangle2D).getBounds2D();
+    }
+
+}
@@ -0,0 +1,190 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.factory;
+
+import static com.iqser.red.service.redaction.v1.server.layoutparsing.document.factory.DocumentGraphFactory.pageBlockIsHeadline;
+import static java.util.Collections.emptyList;
+import static java.util.stream.Collectors.groupingBy;
+
+import java.util.HashSet;
+import java.util.LinkedList;
+import java.util.List;
+import java.util.Map;
+import java.util.Set;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.AbstractPageBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.image.ClassifiedImage;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table.TablePageBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.ClassificationTextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.NodeType;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Page;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Section;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.SemanticNode;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.utils.TableMergingUtility;
+
+import lombok.experimental.UtilityClass;
+
+@UtilityClass
+public class SectionNodeFactory {
+
+    public void addSection(SemanticNode parentNode, List<AbstractPageBlock> pageBlocks, List<ClassifiedImage> images, DocumentGraphFactory.Context context) {
+
+        if (pageBlocks.isEmpty()) {
+            return;
+        }
+        Map<Integer, List<AbstractPageBlock>> blocksPerPage = pageBlocks.stream().collect(groupingBy(AbstractPageBlock::getPage));
+        Section section = Section.builder().entities(new HashSet<>()).documentTree(context.getDocumentTree()).build();
+
+        context.getSections().add(section);
+        blocksPerPage.keySet().forEach(pageNumber -> addSectionNodeToPageNode(context, section, pageNumber));
+
+        List<Integer> tocId;
+        if (parentNode == null) {
+            tocId = context.getDocumentTree().createNewMainEntryAndReturnId(NodeType.SECTION, section);
+        } else {
+            tocId = context.getDocumentTree().createNewChildEntryAndReturnId(parentNode.getTreeId(), NodeType.SECTION, section);
+        }
+        section.setTreeId(tocId);
+
+        addFirstHeadlineDirectlyToSection(pageBlocks, context, section);
+        if (containsTablesAndTextBlocks(pageBlocks)) {
+            splitPageBlocksIntoSubSections(pageBlocks).forEach(subSectionPageBlocks -> addSection(section, subSectionPageBlocks, emptyList(), context));
+        } else {
+            addTablesOrParagraphsOrHeadlinesToSection(pageBlocks, context, section);
+            if (listIsTextBlockOnly(pageBlocks)) {
+                section.setExcludesTables(true);
+            }
+        }
+
+        images.stream().distinct().forEach(image -> DocumentGraphFactory.addImage(section, image, context));
+    }
+
+
+    private static boolean listIsTextBlockOnly(List<AbstractPageBlock> pageBlocks) {
+
+        return pageBlocks.stream().allMatch(abstractPageBlock -> abstractPageBlock instanceof ClassificationTextBlock);
+    }
+
+
+    private void addFirstHeadlineDirectlyToSection(List<AbstractPageBlock> pageBlocks, DocumentGraphFactory.Context context, Section section) {
+
+        if (pageBlockIsHeadline(pageBlocks.get(0))) {
+            addTablesOrParagraphsOrHeadlinesToSection(List.of(pageBlocks.get(0)), context, section);
+            pageBlocks.remove(0);
+        }
+    }
+
+
+    private void addTablesOrParagraphsOrHeadlinesToSection(List<AbstractPageBlock> pageBlocks, DocumentGraphFactory.Context context, Section section) {
+
+        Set<AbstractPageBlock> alreadyMerged = new HashSet<>();
+        List<AbstractPageBlock> remainingBlocks = new LinkedList<>(pageBlocks);
+        for (AbstractPageBlock abstractPageBlock : pageBlocks) {
+
+            if (alreadyMerged.contains(abstractPageBlock)) {
+                continue;
+            }
+
+            remainingBlocks.removeAll(alreadyMerged);
+
+            if (abstractPageBlock instanceof ClassificationTextBlock) {
+                List<ClassificationTextBlock> textBlocks = findTextBlocksWithSameClassificationAndAlignsY(abstractPageBlock, remainingBlocks);
+                alreadyMerged.addAll(textBlocks);
+                DocumentGraphFactory.addParagraphOrHeadline(section, (ClassificationTextBlock) abstractPageBlock, context, textBlocks);
+            }
+            if (abstractPageBlock instanceof TablePageBlock tablePageBlock) {
+
+                List<TablePageBlock> tablesToMerge = TableMergingUtility.findConsecutiveTablesWithSameColCountAndSameHeaders(tablePageBlock, remainingBlocks);
+                alreadyMerged.addAll(tablesToMerge);
+                TableNodeFactory.addTable(section, tablesToMerge, context);
+            }
+        }
+    }
+
+
+    private static boolean containsTablesAndTextBlocks(List<AbstractPageBlock> pageBlocks) {
+
+        return pageBlocks.stream().anyMatch(pageBlock -> pageBlock instanceof TablePageBlock) && pageBlocks.stream()
+                .anyMatch(pageBlock -> pageBlock instanceof ClassificationTextBlock);
+    }
+
+
+    /**
+     * This function splits the list of PageBlocks around TablePageBlocks, such that SubSections can be created, that don't include tables.
+     * First, the function splits the list into coherent Lists, which include only either type.
+     * If a Table is directly preceded by a Headline, it is moved to the Table list.
+     *
+     * @param pageBlocks a List of AbstractPageBlocks, which have at least one TablePageBlock and one ClassificationTextBlock
+     * @return List of Lists of AbstractPageBlocks, which include either a single Headline ClassificationTextBlock and a TablePageBlock or only ClassificationTextBlocks.
+     */
+    private List<List<AbstractPageBlock>> splitPageBlocksIntoSubSections(List<AbstractPageBlock> pageBlocks) {
+
+        List<List<AbstractPageBlock>> splitList = splitIntoCoherentList(pageBlocks);
+        movePrecedingHeadlineToTableList(splitList);
+        return splitList.stream().filter(list -> !list.isEmpty()).toList();
+    }
+
+
+    private static void movePrecedingHeadlineToTableList(List<List<AbstractPageBlock>> splitList) {
+
+        for (int i = 0; i < splitList.size(); i++) {
+            if (listIsTableOnly(splitList.get(i)) && i > 0) {
+                List<AbstractPageBlock> previousList = splitList.get(i - 1);
+                AbstractPageBlock lastPageBlockInPreviousList = previousList.get(previousList.size() - 1);
+                if (pageBlockIsHeadline(lastPageBlockInPreviousList)) {
+                    previousList.remove(i - 1);
+                    splitList.get(i).add(0, lastPageBlockInPreviousList);
+                }
+            }
+        }
+    }
+
+
+    private static boolean listIsTableOnly(List<AbstractPageBlock> abstractPageBlocks) {
+
+        return abstractPageBlocks.stream().allMatch(abstractPageBlock -> abstractPageBlock instanceof TablePageBlock);
+    }
+
+
+    /**
+     * @param pageBlocks a List of AbstractPageBlocks, which have at least one TablePageBlock and one ClassificationTextBlock
+     * @return List of Lists of AbstractPageBlocks, which are exclusively of type ClassificationTextBlock or TablePageBlock
+     */
+    private static List<List<AbstractPageBlock>> splitIntoCoherentList(List<AbstractPageBlock> pageBlocks) {
+
+        List<List<AbstractPageBlock>> splitList = new LinkedList<>();
+        List<AbstractPageBlock> currentList = new LinkedList<>();
+        splitList.add(currentList);
+
+        Class<? extends AbstractPageBlock> lastPageBlockClass = pageBlocks.get(0).getClass();
+        for (AbstractPageBlock pageBlock : pageBlocks) {
+            if (lastPageBlockClass.isInstance(pageBlock)) {
+                currentList.add(pageBlock);
+            } else {
+                currentList = new LinkedList<>();
+                currentList.add(pageBlock);
+                splitList.add(currentList);
+                lastPageBlockClass = pageBlock.getClass();
+            }
+        }
+        return splitList;
+    }
+
+
+    private static List<ClassificationTextBlock> findTextBlocksWithSameClassificationAndAlignsY(AbstractPageBlock atc, List<AbstractPageBlock> pageBlocks) {
+
+        return pageBlocks.stream()
+                .filter(abstractTextContainer -> !abstractTextContainer.equals(atc))
+                .filter(abstractTextContainer -> abstractTextContainer.getPage() == atc.getPage())
+                .filter(abstractTextContainer -> abstractTextContainer instanceof ClassificationTextBlock)
+                .filter(abstractTextContainer -> abstractTextContainer.intersectsY(atc))
+                .map(abstractTextContainer -> (ClassificationTextBlock) abstractTextContainer)
+                .toList();
+    }
+
+
+    private void addSectionNodeToPageNode(DocumentGraphFactory.Context context, Section section, Integer pageNumber) {
+
+        Page page = context.getPage(pageNumber);
+        page.getMainBody().add(section);
+    }
+
+}
@@ -0,0 +1,136 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.factory;
+
+import static java.util.Collections.emptyList;
+
+import java.util.Collection;
+import java.util.List;
+import java.util.Set;
+import java.util.stream.Collectors;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.AbstractPageBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table.Cell;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.table.TablePageBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.TextPositionSequence;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.NodeType;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Page;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.SemanticNode;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Table;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.TableCell;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.utils.TextPositionOperations;
+
+import lombok.experimental.UtilityClass;
+
+@UtilityClass
+public class TableNodeFactory {
+
+    public static final double TABLE_CELL_MERGE_CONTENTS_SIZE_THRESHOLD = 0.05;
+
+
+    public void addTable(SemanticNode parentNode, List<TablePageBlock> tablesToMerge, DocumentGraphFactory.Context context) {
+
+        setPageNumberInCellTextBlocks(tablesToMerge);
+        Set<Page> pages = tablesToMerge.stream().map(AbstractPageBlock::getPage).map(context::getPage).collect(Collectors.toSet());
+        List<List<Cell>> mergedRows = tablesToMerge.stream().map(TablePageBlock::getRows).flatMap(Collection::stream).toList();
+        Table table = Table.builder().documentTree(context.getDocumentTree()).numberOfCols(mergedRows.get(0).size()).numberOfRows(mergedRows.size()).build();
+
+        pages.forEach(page -> addTableToPage(page, parentNode, table));
+
+        List<Integer> tocId = context.getDocumentTree().createNewChildEntryAndReturnId(parentNode.getTreeId(), NodeType.TABLE, table);
+        table.setTreeId(tocId);
+        addTableCells(mergedRows, table, context);
+
+        IfTableHasNoHeadersAssumeFirstRowAreHeaders(table);
+    }
+
+
+    private static void setPageNumberInCellTextBlocks(List<TablePageBlock> tablesToMerge) {
+
+        tablesToMerge.forEach(table -> table.getRows()
+                .stream()
+                .flatMap(Collection::stream)
+                .peek(cell -> cell.setPageNumber(table.getPage()))
+                .forEach(cell -> cell.getTextBlocks().stream().filter(tb -> tb.getPage() == 0).forEach(tb -> tb.setPage(table.getPage()))));
+    }
+
+
+    private static void addTableToPage(Page page, SemanticNode parentNode, Table table) {
+
+        if (!page.getMainBody().contains(parentNode)) {
+            parentNode.getPages().add(page);
+        }
+
+        page.getMainBody().add(table);
+    }
+
+
+    private static void IfTableHasNoHeadersAssumeFirstRowAreHeaders(Table table) {
+
+        if (table.streamHeaders().findAny().isEmpty()) {
+            table.streamRow(0).forEach(tableCellNode -> tableCellNode.setHeader(true));
+        }
+    }
+
+
+    private void addTableCells(List<List<Cell>> rows, Table table, DocumentGraphFactory.Context context) {
+
+        for (int rowIndex = 0; rowIndex < rows.size(); rowIndex++) {
+            for (int colIndex = 0; colIndex < rows.get(rowIndex).size(); colIndex++) {
+                addTableCell(rows.get(rowIndex).get(colIndex), rowIndex, colIndex, table, context);
+            }
+        }
+    }
+
+
+    @SuppressWarnings("PMD.UnusedPrivateMethod") // PMD actually flags this wrong
+    private void addTableCell(Cell cell, int rowIndex, int colIndex, SemanticNode parentNode, DocumentGraphFactory.Context context) {
+
+        Page page = context.getPage(cell.getPageNumber());
+
+        TableCell tableCell = TableCell.builder().documentTree(context.getDocumentTree()).row(rowIndex).col(colIndex).header(cell.isHeaderCell()).bBox(cell.getBounds2D()).build();
+        page.getMainBody().add(tableCell);
+
+        com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.TextBlock textBlock;
+
+        List<Integer> tocId = context.getDocumentTree().createNewChildEntryAndReturnId(parentNode.getTreeId(), NodeType.TABLE_CELL, tableCell);
+        tableCell.setTreeId(tocId);
+
+        if (cell.getTextBlocks().isEmpty()) {
+            tableCell.setLeafTextBlock(context.getTextBlockFactory().emptyTextBlock(parentNode, context, page));
+            tableCell.setLeaf(true);
+
+        } else if (cell.getTextBlocks().size() == 1) {
+            textBlock = context.getTextBlockFactory().buildAtomicTextBlock(cell.getTextBlocks().get(0).getSequences(), tableCell, context, page);
+            tableCell.setLeafTextBlock(textBlock);
+            tableCell.setLeaf(true);
+
+        } else if (firstTextBlockIsHeadline(cell)) {
+            SectionNodeFactory.addSection(tableCell, cell.getTextBlocks().stream().map(tb -> (AbstractPageBlock) tb).toList(), emptyList(), context);
+            tableCell.setLeaf(false);
+
+        } else if (cellAreaIsSmallerThanPageAreaTimesThreshold(cell, page)) {
+            List<TextPositionSequence> sequences = TextPositionOperations.mergeAndSortTextPositionSequenceByYThenX(cell.getTextBlocks());
+            textBlock = context.getTextBlockFactory().buildAtomicTextBlock(sequences, tableCell, context, page);
+            tableCell.setLeafTextBlock(textBlock);
+            tableCell.setLeaf(true);
+
+        } else {
+            cell.getTextBlocks().forEach(tb -> DocumentGraphFactory.addParagraphOrHeadline(tableCell, tb, context, emptyList()));
+            tableCell.setLeaf(false);
+        }
+
+    }
+
+
+    private static boolean cellAreaIsSmallerThanPageAreaTimesThreshold(Cell cell, Page page) {
+
+        return cell.getArea() < TABLE_CELL_MERGE_CONTENTS_SIZE_THRESHOLD * page.getHeight() * page.getWidth();
+    }
+
+
+    private static boolean firstTextBlockIsHeadline(Cell cell) {
+
+        String classification = cell.getTextBlocks().get(0).getClassification();
+        return classification != null && classification.startsWith("H");
+    }
+
+}
@@ -0,0 +1,74 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.factory;
+
+import java.util.Collections;
+import java.util.List;
+import java.util.concurrent.atomic.AtomicInteger;
+import java.util.concurrent.atomic.AtomicLong;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.classification.model.text.TextPositionSequence;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.Boundary;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Page;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.SemanticNode;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.AtomicTextBlock;
+
+public class TextBlockFactory {
+
+    AtomicInteger stringOffset;
+    AtomicLong textBlockIdx;
+
+
+    public TextBlockFactory() {
+
+        stringOffset = new AtomicInteger();
+        textBlockIdx = new AtomicLong();
+    }
+
+
+    public AtomicTextBlock buildAtomicTextBlock(List<TextPositionSequence> sequences, SemanticNode parent, DocumentGraphFactory.Context context, Page page) {
+
+        Integer numberOnPage = context.getPages().get(page).getAndIncrement();
+        return buildAtomicTextBlock(sequences, parent, context, numberOnPage, page);
+    }
+
+
+    public AtomicTextBlock buildAtomicTextBlock(List<TextPositionSequence> sequences, SemanticNode parent, DocumentGraphFactory.Context context, Integer numberOnPage, Page page) {
+
+        SearchTextWithTextPositionDto searchTextWithTextPositionDto = SearchTextWithTextPositionFactory.buildSearchTextToTextPositionModel(sequences);
+        int offset = stringOffset.getAndAdd(searchTextWithTextPositionDto.getSearchText().length());
+
+        return AtomicTextBlock.builder()
+                .id(textBlockIdx.getAndIncrement())
+                .parent(parent)
+                .searchText(searchTextWithTextPositionDto.getSearchText())
+                .numberOnPage(numberOnPage)
+                .page(page)
+                .lineBreaks(searchTextWithTextPositionDto.getLineBreaks())
+                .positions(searchTextWithTextPositionDto.getPositions())
+                .stringIdxToPositionIdx(searchTextWithTextPositionDto.getStringCoordsToPositionCoords())
+                .boundary(new Boundary(offset, offset + searchTextWithTextPositionDto.getSearchText().length()))
+                .build();
+    }
+
+
+    public AtomicTextBlock emptyTextBlock(SemanticNode parent, DocumentGraphFactory.Context context, Page page) {
+
+        return emptyTextBlock(parent, context.getPages().get(page).getAndIncrement(), page);
+    }
+
+
+    public AtomicTextBlock emptyTextBlock(SemanticNode parent, Integer numberOnPage, Page page) {
+
+        return AtomicTextBlock.builder()
+                .id(textBlockIdx.getAndIncrement())
+                .boundary(new Boundary(stringOffset.get(), stringOffset.get()))
+                .searchText("")
+                .lineBreaks(Collections.emptyList())
+                .page(page)
+                .numberOnPage(numberOnPage)
+                .stringIdxToPositionIdx(Collections.emptyList())
+                .positions(Collections.emptyList())
+                .parent(parent)
+                .build();
+    }
+
+}
@@ -0,0 +1,151 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph;
+
+import static java.lang.String.format;
+
+import java.util.Collection;
+import java.util.LinkedList;
+import java.util.List;
+
+import lombok.Setter;
+
+@Setter
+public class Boundary implements Comparable<Boundary> {
+
+    private int start;
+    private int end;
+
+
+    public Boundary(int start, int end) {
+
+        if (start > end) {
+            throw new IllegalArgumentException(format("start: %d > end: %d", start, end));
+        }
+        this.start = start;
+        this.end = end;
+    }
+
+
+    public int length() {
+
+        return end - start;
+    }
+
+
+    public int start() {
+
+        return start;
+    }
+
+
+    public int end() {
+
+        return end;
+    }
+
+
+    public boolean contains(Boundary boundary) {
+
+        return start <= boundary.start() && boundary.end() <= end;
+    }
+
+
+    public boolean containedBy(Boundary boundary) {
+
+        return boundary.contains(this);
+    }
+
+
+    public boolean contains(int start, int end) {
+
+        if (start > end) {
+            throw new IllegalArgumentException(format("start: %d > end: %d", start, end));
+        }
+        return this.start <= start && end <= this.end;
+    }
+
+
+    public boolean containedBy(int start, int end) {
+
+        if (start > end) {
+            throw new IllegalArgumentException(format("start: %d > end: %d", start, end));
+        }
+        return start <= this.start && this.end <= end;
+    }
+
+
+    public boolean contains(int index) {
+
+        return start <= index && index < end;
+    }
+
+
+    public boolean intersects(Boundary boundary) {
+
+        return contains(boundary.start()) || contains(boundary.end() - 1);
+    }
+
+
+    public List<Boundary> split(List<Integer> splitIndices) {
+
+        if (splitIndices.stream().anyMatch(idx -> !this.contains(idx))) {
+            throw new IndexOutOfBoundsException(format("%s splitting indices are out of range for %s", splitIndices.stream().filter(idx -> !this.contains(idx)).toList(), this));
+        }
+        List<Boundary> splitBoundaries = new LinkedList<>();
+        int previousIndex = start;
+        for (int splitIndex : splitIndices) {
+
+            // skip split if it would produce a boundary of length 0
+            if (splitIndex == previousIndex) {
+                continue;
+            }
+            splitBoundaries.add(new Boundary(previousIndex, splitIndex));
+            previousIndex = splitIndex;
+        }
+        splitBoundaries.add(new Boundary(previousIndex, end));
+        return splitBoundaries;
+    }
+
+
+    public static Boundary merge(Collection<Boundary> boundaries) {
+
+        int minStart = boundaries.stream().mapToInt(Boundary::start).min().orElseThrow(IllegalArgumentException::new);
+        int maxEnd = boundaries.stream().mapToInt(Boundary::end).max().orElseThrow(IllegalArgumentException::new);
+        return new Boundary(minStart, maxEnd);
+    }
+
+
+    @Override
+    public String toString() {
+
+        return format("Boundary [%d|%d)", start, end);
+    }
+
+
+    @Override
+    public int compareTo(Boundary boundary) {
+
+        if (end < boundary.end() && start < boundary.start()) {
+            return -1;
+        }
+        if (start > boundary.start() && end > boundary.end()) {
+            return 1;
+        }
+
+        return 0;
+    }
+
+
+    @Override
+    public int hashCode() {
+
+        return toString().hashCode();
+    }
+
+
+    @Override
+    public boolean equals(Object object) {
+
+        return hashCode() == object.hashCode();
+    }
+
+}
@@ -0,0 +1,197 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph;
+
+import static java.lang.String.format;
+
+import java.nio.charset.StandardCharsets;
+import java.util.Collections;
+import java.util.LinkedList;
+import java.util.List;
+import java.util.stream.Stream;
+
+import com.google.common.hash.Hashing;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Document;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.NodeType;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.SemanticNode;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.TextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.TextBlockCollector;
+
+import lombok.AccessLevel;
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.Data;
+import lombok.Getter;
+import lombok.experimental.FieldDefaults;
+
+@Data
+public class DocumentTree {
+
+    private final Entry root;
+
+
+    public DocumentTree(Document document) {
+
+        root = Entry.builder().tocId(Collections.emptyList()).type(NodeType.DOCUMENT).children(new LinkedList<>()).node(document).build();
+    }
+
+
+    public TextBlock buildTextBlock() {
+
+        return streamAllEntriesInOrder().map(Entry::getNode).filter(SemanticNode::isLeaf).map(SemanticNode::getLeafTextBlock).collect(new TextBlockCollector());
+    }
+
+
+    public List<Integer> createNewMainEntryAndReturnId(NodeType nodeType, SemanticNode node) {
+
+        return createNewChildEntryAndReturnId(Collections.emptyList(), nodeType, node);
+    }
+
+
+    public List<Integer> createNewChildEntryAndReturnId(List<Integer> parentId, NodeType nodeType, SemanticNode node) {
+
+        if (!entryExists(parentId)) {
+            throw new UnsupportedOperationException(format("parentId %s does not exist!", parentId));
+        }
+
+        Entry parent = getEntryById(parentId);
+        List<Integer> newId = new LinkedList<>(parentId);
+        newId.add(parent.children.size());
+        parent.children.add(Entry.builder().tocId(newId).node(node).type(nodeType).children(new LinkedList<>()).build());
+
+        return newId;
+    }
+
+
+    private boolean entryExists(List<Integer> tocId) {
+
+        if (tocId.isEmpty()) {
+            return root != null;
+        }
+        Entry entry = root.children.get(tocId.get(0));
+        for (int id : tocId.subList(1, tocId.size())) {
+            if (id >= entry.children.size() || 0 > id) {
+                return false;
+            }
+            entry = entry.children.get(id);
+        }
+        return true;
+    }
+
+
+    public Entry getParentEntryById(List<Integer> tocId) {
+
+        return getEntryById(getParentId(tocId));
+    }
+
+
+    public boolean hasParentById(List<Integer> tocId) {
+
+        return !tocId.isEmpty();
+    }
+
+
+    public Stream<SemanticNode> streamChildrenNodes(List<Integer> tocId) {
+
+        return getEntryById(tocId).children.stream().map(Entry::getNode);
+    }
+
+
+    private static List<Integer> getParentId(List<Integer> tocId) {
+
+        if (tocId.isEmpty()) {
+            throw new UnsupportedOperationException("Root has no parent!");
+        }
+        if (tocId.size() < 2) {
+            return Collections.emptyList();
+        }
+        return tocId.subList(0, tocId.size() - 1);
+    }
+
+
+    public Entry getEntryById(List<Integer> tocId) {
+
+        if (tocId.isEmpty()) {
+            return root;
+        }
+        Entry entry = root.children.get(tocId.get(0));
+        for (int id : tocId.subList(1, tocId.size())) {
+            entry = entry.children.get(id);
+        }
+        return entry;
+    }
+
+
+    public Stream<Entry> streamMainEntries() {
+
+        return root.children.stream();
+    }
+
+
+    public Stream<Entry> streamAllEntriesInOrder() {
+
+        return Stream.of(root).flatMap(DocumentTree::flatten);
+    }
+
+
+    public Stream<Entry> streamAllSubEntriesInOrder(List<Integer> parentId) {
+
+        return getEntryById(parentId).children.stream().flatMap(DocumentTree::flatten);
+    }
+
+
+    @Override
+    public String toString() {
+
+        return String.join("\n", streamAllEntriesInOrder().map(Entry::toString).toList());
+    }
+
+
+    private static Stream<Entry> flatten(Entry entry) {
+
+        return Stream.concat(Stream.of(entry), entry.children.stream().flatMap(DocumentTree::flatten));
+    }
+
+
+    public SemanticNode getHighestParentById(List<Integer> tocId) {
+
+        if (tocId.isEmpty()) {
+            return root.node;
+        }
+        return root.children.get(tocId.get(0)).node;
+    }
+
+
+    @Builder
+    @Getter
+    @AllArgsConstructor
+    @FieldDefaults(level = AccessLevel.PRIVATE, makeFinal = true)
+    public static class Entry {
+
+        List<Integer> tocId;
+        NodeType type;
+        SemanticNode node;
+        List<Entry> children;
+
+
+        @Override
+        public String toString() {
+
+            return node.toString();
+        }
+
+
+        @Override
+        public int hashCode() {
+
+            return Hashing.murmur3_32_fixed().hashString(toString(), StandardCharsets.UTF_8).hashCode();
+        }
+
+
+        @Override
+        public boolean equals(Object o) {
+
+            return o instanceof Entry && o.hashCode() == this.hashCode();
+        }
+
+    }
+
+}
@@ -1,4 +1,4 @@
-package com.iqser.red.service.redaction.v1.server.redaction.model;
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.entity;

 public enum EntityType {
    ENTITY,
@@ -0,0 +1,248 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.entity;
+
+import java.awt.geom.Rectangle2D;
+import java.nio.charset.StandardCharsets;
+import java.util.Collection;
+import java.util.Comparator;
+import java.util.Deque;
+import java.util.HashSet;
+import java.util.LinkedList;
+import java.util.List;
+import java.util.Map;
+import java.util.Set;
+
+import com.google.common.hash.Hashing;
+import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.Engine;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.Boundary;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Page;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.SemanticNode;
+import com.iqser.red.service.redaction.v1.server.redaction.utils.IdBuilder;
+
+import lombok.AccessLevel;
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.Data;
+import lombok.experimental.FieldDefaults;
+
+@Data
+@Builder
+@AllArgsConstructor
+@FieldDefaults(level = AccessLevel.PRIVATE)
+public class RedactionEntity {
+
+    // initial values
+    final Boundary boundary;
+    final String type;
+    final EntityType entityType;
+
+    // empty defaults
+    boolean redaction;
+    boolean removed;
+    boolean ignored;
+    boolean resized;
+    boolean skipRemoveEntitiesContainedInLarger;
+    boolean dictionaryEntry;
+    boolean dossierDictionaryEntry;
+    Set<Engine> engines;
+    Set<RedactionEntity> references;
+    @Builder.Default
+    Deque<Integer> matchedRules = new LinkedList<>();
+    String redactionReason;
+    String legalBasis;
+
+    // inferred on graph insertion
+    String value;
+    String textBefore;
+    String textAfter;
+    @Builder.Default
+    Set<Page> pages = new HashSet<>();
+    List<RedactionPosition> redactionPositionsPerPage;
+    @Builder.Default
+    List<SemanticNode> intersectingNodes = new LinkedList<>();
+    SemanticNode deepestFullyContainingNode;
+
+
+    public static RedactionEntity initialEntityNode(Boundary boundary, String type, EntityType entityType) {
+
+        return RedactionEntity.builder()
+                .type(type)
+                .entityType(entityType)
+                .boundary(boundary)
+                .redaction(false)
+                .removed(false)
+                .ignored(false)
+                .resized(false)
+                .skipRemoveEntitiesContainedInLarger(false)
+                .dictionaryEntry(false)
+                .dossierDictionaryEntry(false)
+                .engines(new HashSet<>())
+                .references(new HashSet<>())
+                .redactionReason("")
+                .legalBasis("")
+                .build();
+    }
+
+
+    public boolean occursInNodeOfType(Class<? extends SemanticNode> clazz) {
+
+        return intersectingNodes.stream().anyMatch(clazz::isInstance);
+    }
+
+
+    public boolean isType(String type) {
+
+        return this.type.equals(type);
+    }
+
+
+    public boolean isAnyType(List<String> types) {
+
+        return types.contains(type);
+    }
+
+
+    public void addIntersectingNode(SemanticNode containingNode) {
+
+        intersectingNodes.add(containingNode);
+    }
+
+
+    public void removeFromGraph() {
+
+        intersectingNodes.forEach(node -> node.getEntities().remove(this));
+        pages.forEach(page -> page.getEntities().remove(this));
+        intersectingNodes = new LinkedList<>();
+        deepestFullyContainingNode = null;
+        pages = new HashSet<>();
+        removed = true;
+        ignored = true;
+    }
+
+
+    public void addMatchedRule(int ruleNumber) {
+
+        matchedRules.add(ruleNumber);
+    }
+
+
+    public int getMatchedRule() {
+
+        if (matchedRules.isEmpty()) {
+            return -1;
+        }
+        return matchedRules.getLast();
+    }
+
+
+    public List<RedactionPosition> getRedactionPositionsPerPage() {
+
+        if (redactionPositionsPerPage == null || redactionPositionsPerPage.isEmpty()) {
+            Map<Page, List<Rectangle2D>> rectanglesPerLinePerPage = deepestFullyContainingNode.buildTextBlock().getPositionsPerPage(boundary);
+
+            Page firstPage = rectanglesPerLinePerPage.keySet()
+                    .stream()
+                    .min(Comparator.comparingInt(Page::getNumber))
+                    .orElseThrow(() -> new RuntimeException("No Positions found on any page!"));
+
+            String id = IdBuilder.buildId(pages, rectanglesPerLinePerPage.values().stream().flatMap(Collection::stream).toList());
+            redactionPositionsPerPage = rectanglesPerLinePerPage.entrySet().stream().map(entry -> buildRedactionPosition(firstPage, id, entry)).toList();
+        }
+        return redactionPositionsPerPage;
+    }
+
+
+    private static RedactionPosition buildRedactionPosition(Page firstPage, String id, Map.Entry<Page, List<Rectangle2D>> entry) {
+
+        if (entry.getKey().equals(firstPage)) {
+            return new RedactionPosition(id, entry.getKey(), entry.getValue());
+        } else {
+            return new RedactionPosition(id + "-" + entry.getKey().getNumber(), entry.getKey(), entry.getValue());
+        }
+    }
+
+
+    public boolean containedBy(RedactionEntity redactionEntity) {
+
+        return this.boundary.containedBy(redactionEntity.getBoundary());
+    }
+
+
+    public boolean contains(RedactionEntity redactionEntity) {
+
+        return this.boundary.contains(redactionEntity.getBoundary());
+    }
+
+
+    public boolean intersects(RedactionEntity redactionEntity) {
+
+        return this.boundary.intersects(redactionEntity.getBoundary());
+    }
+
+
+    public void addEngine(Engine engine) {
+
+        engines.add(engine);
+    }
+
+
+    public void addEngines(Set<Engine> engines) {
+
+        this.engines.addAll(engines);
+    }
+
+
+    public void addReference(RedactionEntity reference) {
+
+        references.add(reference);
+    }
+
+
+    public void addReferences(List<RedactionEntity> references) {
+
+        this.references.addAll(references);
+    }
+
+
+    public boolean matchesAnnotationId(String manualRedactionId) {
+
+        return getRedactionPositionsPerPage().stream().anyMatch(entityPosition -> entityPosition.getId().equals(manualRedactionId));
+    }
+
+
+    @Override
+    public String toString() {
+
+        StringBuilder sb = new StringBuilder();
+        sb.append("Entity[\"");
+        sb.append(value);
+        sb.append("\", ");
+        sb.append(boundary);
+        sb.append(", pages[");
+        pages.forEach(page -> {
+            sb.append(page.getNumber());
+            sb.append(", ");
+        });
+        sb.delete(sb.length() - 2, sb.length());
+        sb.append("], type = \"");
+        sb.append(type);
+        sb.append("\", EntityType.");
+        sb.append(entityType);
+        sb.append("]");
+        return sb.toString();
+    }
+
+
+    @Override
+    public int hashCode() {
+
+        return Hashing.murmur3_128().hashString(toString(), StandardCharsets.UTF_8).hashCode();
+    }
+
+
+    @Override
+    public boolean equals(Object o) {
+
+        return o.hashCode() == hashCode();
+    }
+
+}
@@ -0,0 +1,22 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.entity;
+
+import java.awt.geom.Rectangle2D;
+import java.util.List;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes.Page;
+
+import lombok.AccessLevel;
+import lombok.AllArgsConstructor;
+import lombok.Data;
+import lombok.experimental.FieldDefaults;
+
+@Data
+@AllArgsConstructor
+@FieldDefaults(level = AccessLevel.PRIVATE)
+public class RedactionPosition {
+
+    final String id;
+    Page page;
+    List<Rectangle2D> rectanglePerLine;
+
+}
@@ -0,0 +1,112 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes;
+
+import java.awt.geom.Rectangle2D;
+import java.util.Collections;
+import java.util.HashMap;
+import java.util.HashSet;
+import java.util.List;
+import java.util.Map;
+import java.util.Set;
+import java.util.stream.Collectors;
+import java.util.stream.Stream;
+
+import com.iqser.red.service.redaction.v1.server.exception.NotFoundException;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.DocumentTree;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.entity.RedactionEntity;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.TextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.TextBlockCollector;
+
+import lombok.AccessLevel;
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.Data;
+import lombok.NoArgsConstructor;
+import lombok.experimental.FieldDefaults;
+
+@Data
+@Builder
+@AllArgsConstructor
+@NoArgsConstructor
+@FieldDefaults(level = AccessLevel.PRIVATE)
+public class Document implements SemanticNode {
+
+    Set<Page> pages;
+    DocumentTree documentTree;
+    Integer numberOfPages;
+    TextBlock textBlock;
+    @Builder.Default
+    Set<RedactionEntity> entities = new HashSet<>();
+
+
+    public TextBlock buildTextBlock() {
+
+        return streamTerminalTextBlocksInOrder().collect(new TextBlockCollector());
+    }
+
+
+    public List<Section> getMainSections() {
+
+        return streamChildren().filter(node -> node instanceof Section).map(node -> (Section) node).collect(Collectors.toList());
+    }
+
+
+    public Stream<TextBlock> streamTerminalTextBlocksInOrder() {
+
+        return streamAllNodes().filter(SemanticNode::isLeaf).map(SemanticNode::getLeafTextBlock);
+    }
+
+
+    @Override
+    public List<Integer> getTreeId() {
+
+        return Collections.emptyList();
+    }
+
+
+    @Override
+    public void setTreeId(List<Integer> tocId) {
+
+        throw new UnsupportedOperationException("Document is always the root of the TablePageBlock of Contents");
+    }
+
+
+    @Override
+    public Headline getHeadline() {
+
+        return streamAllSubNodes().filter(node -> node instanceof Headline)
+                .map(node -> (Headline) node)
+                .findFirst()
+                .orElseThrow(() -> new NotFoundException("No Headlines found in this document!"));
+    }
+
+
+    private Stream<SemanticNode> streamAllNodes() {
+
+        return documentTree.streamAllEntriesInOrder().map(DocumentTree.Entry::getNode);
+    }
+
+
+    public Stream<Image> streamAllImages() {
+
+        return streamAllSubNodes().filter(node -> node instanceof Image).map(node -> (Image) node);
+    }
+
+
+    @Override
+    public String toString() {
+
+        return NodeType.DOCUMENT + ": " + buildTextBlock().buildSummary();
+    }
+
+
+    @Override
+    public Map<Page, Rectangle2D> getBBox() {
+
+        Map<Page, Rectangle2D> bBox = new HashMap<>();
+        for (Page page : pages) {
+            bBox.put(page, new Rectangle2D.Double(0, 0, page.getWidth(), page.getHeight()));
+        }
+        return bBox;
+    }
+
+}
@@ -0,0 +1,53 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes;
+
+import java.util.HashSet;
+import java.util.List;
+import java.util.Set;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.DocumentTree;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.entity.RedactionEntity;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.TextBlock;
+
+import lombok.AccessLevel;
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.Data;
+import lombok.EqualsAndHashCode;
+import lombok.NoArgsConstructor;
+import lombok.experimental.FieldDefaults;
+
+@Data
+@Builder
+@AllArgsConstructor
+@NoArgsConstructor
+@FieldDefaults(level = AccessLevel.PRIVATE)
+public class Footer implements SemanticNode {
+
+    List<Integer> treeId;
+    TextBlock leafTextBlock;
+
+    @Builder.Default
+    boolean leaf = true;
+
+    @EqualsAndHashCode.Exclude
+    DocumentTree documentTree;
+
+    @Builder.Default
+    @EqualsAndHashCode.Exclude
+    Set<RedactionEntity> entities = new HashSet<>();
+
+
+    @Override
+    public TextBlock buildTextBlock() {
+
+        return leafTextBlock;
+    }
+
+
+    @Override
+    public String toString() {
+
+        return treeId + ": " + NodeType.FOOTER + ": " + leafTextBlock.buildSummary();
+    }
+
+}
@@ -0,0 +1,53 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes;
+
+import java.util.HashSet;
+import java.util.List;
+import java.util.Set;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.DocumentTree;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.entity.RedactionEntity;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.TextBlock;
+
+import lombok.AccessLevel;
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.Data;
+import lombok.EqualsAndHashCode;
+import lombok.NoArgsConstructor;
+import lombok.experimental.FieldDefaults;
+
+@Data
+@Builder
+@AllArgsConstructor
+@NoArgsConstructor
+@FieldDefaults(level = AccessLevel.PRIVATE)
+public class Header implements SemanticNode {
+
+    List<Integer> treeId;
+    TextBlock leafTextBlock;
+
+    @Builder.Default
+    boolean leaf = true;
+
+    @EqualsAndHashCode.Exclude
+    DocumentTree documentTree;
+
+    @Builder.Default
+    @EqualsAndHashCode.Exclude
+    Set<RedactionEntity> entities = new HashSet<>();
+
+
+    @Override
+    public TextBlock buildTextBlock() {
+
+        return leafTextBlock;
+    }
+
+
+    @Override
+    public String toString() {
+
+        return treeId + ": " + NodeType.HEADER + ": " + leafTextBlock.buildSummary();
+    }
+
+}
@@ -0,0 +1,60 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes;
+
+import java.util.HashSet;
+import java.util.List;
+import java.util.Set;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.DocumentTree;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.entity.RedactionEntity;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.TextBlock;
+
+import lombok.AccessLevel;
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.Data;
+import lombok.EqualsAndHashCode;
+import lombok.NoArgsConstructor;
+import lombok.experimental.FieldDefaults;
+
+@Data
+@Builder
+@AllArgsConstructor
+@NoArgsConstructor
+@FieldDefaults(level = AccessLevel.PRIVATE)
+public class Headline implements SemanticNode {
+
+    List<Integer> treeId;
+    TextBlock leafTextBlock;
+
+    @Builder.Default
+    boolean leaf = true;
+
+    @EqualsAndHashCode.Exclude
+    DocumentTree documentTree;
+
+    @Builder.Default
+    @EqualsAndHashCode.Exclude
+    Set<RedactionEntity> entities = new HashSet<>();
+
+
+    @Override
+    public TextBlock buildTextBlock() {
+
+        return leafTextBlock;
+    }
+
+
+    @Override
+    public String toString() {
+
+        return treeId + ": " + NodeType.HEADLINE + ": " + leafTextBlock.buildSummary();
+    }
+
+
+    @Override
+    public Headline getHeadline() {
+
+        return this;
+    }
+
+}
@@ -0,0 +1,87 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes;
+
+import java.awt.geom.Rectangle2D;
+import java.util.Collections;
+import java.util.HashMap;
+import java.util.HashSet;
+import java.util.List;
+import java.util.Map;
+import java.util.Set;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.DocumentTree;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.entity.RedactionEntity;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.TextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.TextBlockCollector;
+
+import lombok.AccessLevel;
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.Data;
+import lombok.EqualsAndHashCode;
+import lombok.NoArgsConstructor;
+import lombok.experimental.FieldDefaults;
+
+@Data
+@Builder
+@AllArgsConstructor
+@NoArgsConstructor
+@FieldDefaults(level = AccessLevel.PRIVATE)
+public class Image implements SemanticNode {
+
+    List<Integer> treeId;
+    String id;
+
+    ImageType imageType;
+    boolean transparent;
+    Rectangle2D position;
+
+    boolean redaction;
+    boolean ignored;
+    @Builder.Default
+    String redactionReason = "";
+    @Builder.Default
+    String legalBasis = "";
+    @Builder.Default
+    int matchedRule = -1;
+
+    @EqualsAndHashCode.Exclude
+    Page page;
+
+    @EqualsAndHashCode.Exclude
+    DocumentTree documentTree;
+
+    @Builder.Default
+    @EqualsAndHashCode.Exclude
+    Set<RedactionEntity> entities = new HashSet<>();
+
+
+    @Override
+    public TextBlock buildTextBlock() {
+
+        return streamAllSubNodes().filter(SemanticNode::isLeaf).map(SemanticNode::getLeafTextBlock).collect(new TextBlockCollector());
+    }
+
+
+    @Override
+    public Set<Page> getPages() {
+
+        return Collections.singleton(page);
+    }
+
+
+    @Override
+    public String toString() {
+
+        return treeId + ": " + NodeType.IMAGE + ": " + imageType.toString() + " " + position;
+    }
+
+
+    @Override
+    public Map<Page, Rectangle2D> getBBox() {
+
+        Map<Page, Rectangle2D> bBoxPerPage = new HashMap<>();
+        bBoxPerPage.put(page, position);
+        return bBoxPerPage;
+    }
+
+}
@@ -0,0 +1,9 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes;
+
+public enum ImageType {
+    LOGO,
+    FORMULA,
+    SIGNATURE,
+    OTHER,
+    OCR
+}
@@ -0,0 +1,13 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes;
+
+public enum NodeType {
+    DOCUMENT,
+    SECTION,
+    HEADLINE,
+    PARAGRAPH,
+    TABLE,
+    TABLE_CELL,
+    IMAGE,
+    HEADER,
+    FOOTER
+}
@@ -0,0 +1,73 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes;
+
+import java.util.HashSet;
+import java.util.List;
+import java.util.Set;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.entity.RedactionEntity;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.TextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.TextBlockCollector;
+
+import lombok.AccessLevel;
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.EqualsAndHashCode;
+import lombok.Getter;
+import lombok.Setter;
+import lombok.experimental.FieldDefaults;
+
+@Getter
+@Setter
+@Builder
+@AllArgsConstructor
+@FieldDefaults(level = AccessLevel.PRIVATE)
+public class Page {
+
+    Integer number;
+    Integer height;
+    Integer width;
+    Integer rotation;
+
+    @EqualsAndHashCode.Exclude
+    List<SemanticNode> mainBody;
+    @EqualsAndHashCode.Exclude
+    Header header;
+    @EqualsAndHashCode.Exclude
+    Footer footer;
+
+    @Builder.Default
+    @EqualsAndHashCode.Exclude
+    Set<RedactionEntity> entities = new HashSet<>();
+
+    @Builder.Default
+    @EqualsAndHashCode.Exclude
+    Set<Image> images = new HashSet<>();
+
+
+    public TextBlock getMainBodyTextBlock() {
+
+        return mainBody.stream().filter(SemanticNode::isLeaf).map(SemanticNode::getLeafTextBlock).collect(new TextBlockCollector());
+    }
+
+
+    @Override
+    public String toString() {
+
+        return String.valueOf(number);
+    }
+
+
+    @Override
+    public int hashCode() {
+
+        return number;
+    }
+
+
+    @Override
+    public boolean equals(Object o) {
+
+        return o instanceof Page && o.hashCode() == this.hashCode();
+    }
+
+}
@@ -0,0 +1,51 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes;
+
+import java.util.HashSet;
+import java.util.List;
+import java.util.Set;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.DocumentTree;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.entity.RedactionEntity;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.TextBlock;
+
+import lombok.AccessLevel;
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.Data;
+import lombok.EqualsAndHashCode;
+import lombok.experimental.FieldDefaults;
+
+@Data
+@Builder
+@AllArgsConstructor
+@FieldDefaults(level = AccessLevel.PRIVATE)
+public class Paragraph implements SemanticNode {
+
+    List<Integer> treeId;
+    TextBlock leafTextBlock;
+
+    @Builder.Default
+    boolean leaf = true;
+
+    @EqualsAndHashCode.Exclude
+    DocumentTree documentTree;
+
+    @Builder.Default
+    @EqualsAndHashCode.Exclude
+    Set<RedactionEntity> entities = new HashSet<>();
+
+
+    @Override
+    public TextBlock buildTextBlock() {
+
+        return leafTextBlock;
+    }
+
+
+    @Override
+    public String toString() {
+
+        return treeId + ": " + NodeType.PARAGRAPH + ": " + leafTextBlock.buildSummary();
+    }
+
+}
@@ -0,0 +1,65 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes;
+
+import java.util.HashSet;
+import java.util.List;
+import java.util.Set;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.DocumentTree;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.entity.RedactionEntity;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.TextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.TextBlockCollector;
+
+import lombok.AccessLevel;
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.Data;
+import lombok.EqualsAndHashCode;
+import lombok.experimental.FieldDefaults;
+import lombok.extern.slf4j.Slf4j;
+
+@Slf4j
+@Data
+@Builder
+@AllArgsConstructor
+@FieldDefaults(level = AccessLevel.PRIVATE)
+public class Section implements SemanticNode {
+
+    List<Integer> treeId;
+
+    TextBlock textBlock;
+    @EqualsAndHashCode.Exclude
+    DocumentTree documentTree;
+    boolean excludesTables;
+
+    @Builder.Default
+    @EqualsAndHashCode.Exclude
+    Set<RedactionEntity> entities = new HashSet<>();
+
+
+    @Override
+    public TextBlock buildTextBlock() {
+
+        if (textBlock == null) {
+            textBlock = streamAllSubNodes().filter(SemanticNode::isLeaf).map(SemanticNode::getLeafTextBlock).collect(new TextBlockCollector());
+        }
+        return textBlock;
+    }
+
+
+    @Override
+    public String toString() {
+
+        return treeId.toString() + ": " + NodeType.SECTION + ": " + buildTextBlock().buildSummary();
+    }
+
+
+    public Headline getHeadline() {
+
+        return streamChildren()//
+                .filter(node -> node instanceof Headline)//
+                .map(node -> (Headline) node)//
+                .findFirst()//
+                .orElseGet(() -> getParent().getHeadline());
+    }
+
+}
@@ -0,0 +1,407 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes;
+
+import static java.lang.String.format;
+
+import java.awt.geom.Rectangle2D;
+import java.util.HashMap;
+import java.util.List;
+import java.util.Map;
+import java.util.Set;
+import java.util.stream.Collectors;
+import java.util.stream.Stream;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.Boundary;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.DocumentTree;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.entity.RedactionEntity;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.AtomicTextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.TextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.utils.RectangleTransformations;
+
+public interface SemanticNode {
+
+    /**
+     * Searches all Nodes located underneath this Node in the DocumentTree and concatenates their AtomicTextBlocks into a single TextBlockEntity.
+     * So, for a Section all TextBlocks of Subsections, Paragraphs, and Tables are concatenated into a single TextBlockEntity
+     * If the Node is a Leaf, the LeafTextBlock will be returned instead.
+     *
+     * @return TextBlock containing all AtomicTextBlocks that are located under this Node.
+     */
+    TextBlock buildTextBlock();
+
+
+    /**
+     * Any Node maintains its own Set of Entities.
+     * This Set contains all Entities whose boundary intersects the boundary of this node.
+     *
+     * @return Set of all Entities associated with this Node
+     */
+    Set<RedactionEntity> getEntities();
+
+
+    /**
+     * Each AtomicTextBlock is assigned a page, so to get the pages this node appears on, it collects the PageNodes from each AtomicTextBlock belonging to this node's TextBlock.
+     *
+     * @return Set of PageNodes this node appears on.
+     */
+    default Set<Page> getPages() {
+
+        return buildTextBlock().getPages();
+    }
+
+
+    /**
+     * Each AtomicTextBlock is assigned a page, so to get the pages for this boundary, it collects the PageNodes from each AtomicTextBlock belonging to this node's TextBlock.
+     *
+     * @return Set of PageNodes this node appears on.
+     */
+    default Set<Page> getPages(Boundary boundary) {
+
+        if (!getBoundary().contains(boundary)) {
+            throw new IllegalArgumentException(format("%s which was used to query foir pages is not contained in the %s of this node!", getBoundary(), boundary));
+        }
+        return buildTextBlock().getPages(boundary);
+    }
+
+
+    /**
+     * Returns the DocumentTree Object.
+     *
+     * @return the DocumentTree of the Document this node belongs to
+     */
+    DocumentTree getDocumentTree();
+
+
+    /**
+     * The id is a List of Integers uniquely identifying this node in the DocumentTree.
+     *
+     * @return the DocumentTree ID
+     */
+    List<Integer> getTreeId();
+
+
+    /**
+     * This should only be used during graph construction.
+     *
+     * @param tocId List of Integers
+     */
+    void setTreeId(List<Integer> tocId);
+
+
+    /**
+     * Traverses the Tree up, until it hits a Headline or hits a Section which will then return the first Headline from its children.
+     * Throws NotFoundException if no Headline is found this way
+     *
+     * @return First Headline found
+     */
+    default Headline getHeadline() {
+
+        return getParent().getHeadline();
+    }
+
+
+    /**
+     * Checks if its TocId has a length greater than zero.
+     *
+     * @return boolean indicating whether this Node has a Parent in the DocumentTree
+     */
+    default boolean hasParent() {
+
+        return getDocumentTree().hasParentById(getTreeId());
+    }
+
+
+    /**
+     * @return The SemanticNode representing the Parent in the DocumentTree
+     * throws NotFoundException, when no parent is present
+     */
+    default SemanticNode getParent() {
+
+        return getDocumentTree().getParentEntryById(getTreeId()).getNode();
+    }
+
+
+    /**
+     * @return The SemanticNode which is directly underneath the document and also under which this node is.
+     * if this is the highest child node or the document itself, it returns itself.
+     */
+    default SemanticNode getHighestParent() {
+
+        return getDocumentTree().getHighestParentById(getTreeId());
+    }
+
+
+    /**
+     * Leaf means a SemanticNode has direct access to a TextBlock, by default this is false and must be overridden.
+     * Currently only Sections, Images, and Tables are not leaves.
+     * A TableCell might be a leaf depending on its area compared to the page.
+     *
+     * @return boolean, indicating if a Node has direct access to a TextBlock
+     */
+    default boolean isLeaf() {
+
+        return false;
+    }
+
+
+    /**
+     * Leaf means a SemanticNode has direct access to a TextBlock, by default this is false and must be overridden.
+     * Currently only Sections and Tables are no leaves.
+     *
+     * @return AtomicTextBlock
+     */
+    default TextBlock getLeafTextBlock() {
+
+        throw new UnsupportedOperationException("Only leaf Nodes have access to LeafTextBlocks!");
+    }
+
+
+    /**
+     * Should only be used during construction of the Graph. Sets the LeafTextBlock of this SemanticNode.
+     *
+     * @param textBlock the TextBlock to set as the LeafTextBlock of this SemanticNode
+     */
+    default void setLeafTextBlock(TextBlock textBlock) {
+
+        throw new UnsupportedOperationException();
+    }
+
+
+    /**
+     * Checks whether this SemanticNode has any Entity of the provided type.
+     *
+     * @param type string representing the type of entity to check for
+     * @return true, if this SemanticNode has at least one Entity of the provided type
+     */
+    default boolean hasEntitiesOfType(String type) {
+
+        return getEntities().stream().anyMatch(redactionEntity -> redactionEntity.getType().equals(type));
+    }
+
+
+    /**
+     * Returns a List of Entities in this SemanticNode which are of the provided type such as "CBI_author".
+     *
+     * @param type string representing the type of entities to return
+     * @return List of RedactionEntities of any the type
+     */
+    default List<RedactionEntity> getEntitiesOfType(String type) {
+
+        return getEntities().stream().filter(redactionEntity -> redactionEntity.getType().equals(type)).toList();
+    }
+
+
+    /**
+     * Returns a List of Entities in this SemanticNode which have any of the provided types such as "CBI_author".
+     *
+     * @param types A list of strings representing the types of entities to return
+     * @return List of RedactionEntities of any provided type
+     */
+    default List<RedactionEntity> getEntitiesOfType(List<String> types) {
+
+        return getEntities().stream().filter(redactionEntity -> redactionEntity.isAnyType(types)).toList();
+    }
+
+
+    /**
+     * Each AtomicTextBlock has an index on its page, this returns the number of the first AtomicTextBlock underneath this node.
+     * If this node does not have any AtomicTexBlocks underneath it, e.g. an empty TableCell. It returns -1.
+     *
+     * @return Integer representing the number on the page
+     */
+    default Integer getNumberOnPage() {
+
+        TextBlock textBlock = buildTextBlock();
+        if (textBlock.getAtomicTextBlocks().size() > 0) {
+            return buildTextBlock().getAtomicTextBlocks().get(0).getNumberOnPage();
+        } else {
+            return -1;
+        }
+    }
+
+
+    /**
+     * Checks if the SemanticNode contains any text.
+     *
+     * @return true, if this node's TextBlock is not empty
+     */
+    default boolean hasText() {
+
+        return buildTextBlock().length() > 0;
+    }
+
+
+    /**
+     * Checks whether this SemanticNode contains the provided String.
+     *
+     * @param string A String which the TextBlock might contain
+     * @return true, if this node's TextBlock contains the string
+     */
+    default boolean containsString(String string) {
+
+        return buildTextBlock().getSearchText().contains(string);
+    }
+
+
+    /**
+     * Checks whether this SemanticNode contains all the provided Strings.
+     *
+     * @param strings A List of Strings which the TextBlock might contain
+     * @return true, if this node's TextBlock contains all strings
+     */
+    default boolean containsStrings(List<String> strings) {
+
+        return strings.stream().allMatch(this::containsString);
+    }
+
+
+    /**
+     * Checks whether this SemanticNode contains all the provided Strings ignoring case.
+     *
+     * @param string A String which the TextBlock might contain
+     * @return true, if this node's TextBlock contains the string ignoring case
+     */
+    default boolean containsStringIgnoreCase(String string) {
+
+        return buildTextBlock().getSearchText().toLowerCase().contains(string.toLowerCase());
+    }
+
+
+    /**
+     * Checks whether this SemanticNode contains any of the provided Strings.
+     *
+     * @param strings A List of Strings which the TextBlock might contain
+     * @return true, if this node's TextBlock contains any of the strings
+     */
+    default boolean containsAnyString(List<String> strings) {
+
+        return strings.stream().anyMatch(this::containsString);
+    }
+
+
+    /**
+     * Checks whether this SemanticNode contains any of the provided Strings ignoring case.
+     *
+     * @param strings A List of Strings which the TextBlock might contain
+     * @return true, if this node's TextBlock contains any of the strings
+     */
+    default boolean containsAnyStringIgnoreCase(List<String> strings) {
+
+        return strings.stream().anyMatch(this::containsStringIgnoreCase);
+    }
+
+
+    /**
+     * This function is used during insertion of EntityNodes into the graph, it checks if the boundary of the RedactionEntity intersects or even contains the RedactionEntity.
+     * It sets the fields accordingly and recursively calls this function on all its children.
+     *
+     * @param redactionEntity RedactionEntity, which is being inserted into the graph
+     */
+    default void addThisToEntityIfIntersects(RedactionEntity redactionEntity) {
+
+        TextBlock textBlock = buildTextBlock();
+        if (textBlock.getBoundary().intersects(redactionEntity.getBoundary())) {
+            if (textBlock.containsBoundary(redactionEntity.getBoundary())) {
+                redactionEntity.setDeepestFullyContainingNode(this);
+            }
+
+            redactionEntity.addIntersectingNode(this);
+            streamChildren().filter(semanticNode -> semanticNode.getBoundary().intersects(redactionEntity.getBoundary()))
+                    .forEach(node -> node.addThisToEntityIfIntersects(redactionEntity));
+        }
+    }
+
+
+    /**
+     * Streams all children located directly underneath this node in the DocumentTree.
+     *
+     * @return Stream of all children
+     */
+    default Stream<SemanticNode> streamChildren() {
+
+        return getDocumentTree().streamChildrenNodes(getTreeId());
+    }
+
+
+    /**
+     * Recursively streams all SemanticNodes located underneath this node in the DocumentTree in order.
+     *
+     * @return Stream of all SubNodes
+     */
+    default Stream<SemanticNode> streamAllSubNodes() {
+
+        return getDocumentTree().streamAllSubEntriesInOrder(getTreeId()).map(DocumentTree.Entry::getNode);
+    }
+
+
+    /**
+     * The Boundary is the start and end string offsets in the reading order of the document.
+     *
+     * @return Boundary of this Node's TextBlock
+     */
+    default Boundary getBoundary() {
+
+        return buildTextBlock().getBoundary();
+    }
+
+
+    /**
+     * If this Node is a Leaf it will calculate the boundingBox of its LeafTextBlock, otherwise it will calculate the Union of the BoundingBoxes of all its Children.
+     * If called on the Document, it will return the cropbox of each page
+     *
+     * @return Rectangle2D fully encapsulating this Node for each page.
+     */
+    default Map<Page, Rectangle2D> getBBox() {
+
+        Map<Page, Rectangle2D> bBoxPerPage = new HashMap<>();
+        if (isLeaf()) {
+            return getBBoxFromLeafTextBlock(bBoxPerPage);
+        }
+
+        return getBBoxFromChildren(bBoxPerPage);
+    }
+
+
+    /**
+     * Checks whether the Bounding Box of this SemanticNode contains the provided rectangle on the provided page.
+     *
+     * @param rectangle2D The rectangle to check if it is contained
+     * @param pageNumber  The Page number on which the rectangle should be checked
+     * @return boolean
+     */
+    default boolean containsRectangle(Rectangle2D rectangle2D, Integer pageNumber) {
+
+        Page helperPage = Page.builder().number(pageNumber).build();
+        if (!getPages().contains(helperPage)) {
+            return false;
+        }
+        return getBBox().get(helperPage).contains(rectangle2D);
+    }
+
+
+    /**
+     * TODO: this produces unwanted results for sections spanning multiple columns.
+     *
+     * @param bBoxPerPage initial empty BoundingBox
+     * @return The union of the BoundingBoxes of all children
+     */
+    private Map<Page, Rectangle2D> getBBoxFromChildren(Map<Page, Rectangle2D> bBoxPerPage) {
+
+        return streamChildren().map(SemanticNode::getBBox).reduce((map1, map2) -> {
+            map1.forEach((page, rectangle) -> map2.merge(page, rectangle, (rect1, rect2) -> rect1.createUnion(rect2).getBounds2D()));
+            return map2;
+        }).orElse(bBoxPerPage);
+    }
+
+
+    /**
+     * @param bBoxPerPage initial empty BoundingBox
+     * @return The union of all BoundingBoxes of the TextBlock of this node
+     */
+    private Map<Page, Rectangle2D> getBBoxFromLeafTextBlock(Map<Page, Rectangle2D> bBoxPerPage) {
+
+        Map<Page, List<AtomicTextBlock>> atomicTextBlockPerPage = buildTextBlock().getAtomicTextBlocks().stream().collect(Collectors.groupingBy(AtomicTextBlock::getPage));
+        atomicTextBlockPerPage.forEach((page, atbs) -> bBoxPerPage.put(page, RectangleTransformations.bBoxUnionAtomicTextBlock(atbs)));
+        return bBoxPerPage;
+    }
+
+}
@@ -0,0 +1,290 @@
+package com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.nodes;
+
+import static java.lang.String.format;
+
+import java.util.Collection;
+import java.util.HashSet;
+import java.util.List;
+import java.util.Set;
+import java.util.stream.IntStream;
+import java.util.stream.Stream;
+
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.DocumentTree;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.entity.RedactionEntity;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.TextBlock;
+import com.iqser.red.service.redaction.v1.server.layoutparsing.document.graph.textblock.TextBlockCollector;
+
+import lombok.AccessLevel;
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.Data;
+import lombok.EqualsAndHashCode;
+import lombok.experimental.FieldDefaults;
+
+@Data
+@Builder
+@AllArgsConstructor
+@FieldDefaults(level = AccessLevel.PRIVATE)
+public class Table implements SemanticNode {
+
+    List<Integer> treeId;
+    DocumentTree documentTree;
+
+    int numberOfRows;
+    int numberOfCols;
+
+    TextBlock textBlock;
+
+    @Builder.Default
+    @EqualsAndHashCode.Exclude
+    Set<RedactionEntity> entities = new HashSet<>();
+
+
+    /**
+     * Streams all entities in this table, that appear in a row, which contains any of the provided strings.
+     *
+     * @param strings Strings to check whether a row contains them
+     * @return Stream of all entities in this table, that appear in a row, which contains any of the provided strings
+     */
+    public Stream<RedactionEntity> streamEntitiesWhereRowContainsStringsIgnoreCase(List<String> strings) {
+
+        return IntStream.range(0, numberOfRows)
+                .boxed()
+                .filter(row -> rowContainsStringsIgnoreCase(row, strings))
+                .flatMap(this::streamRow)
+                .map(TableCell::getEntities)
+                .flatMap(Collection::stream);
+    }
+
+
+    /**
+     * Checks whether the specified row contains all the provided strings.
+     *
+     * @param row     the row to check as an Integer, must be smaller than numberOfRows
+     * @param strings a list of strings to check for
+     * @return true, if all strings appear in the provided row
+     */
+    public boolean rowContainsStringsIgnoreCase(Integer row, List<String> strings) {
+
+        String rowText = streamRow(row).map(TableCell::buildTextBlock).collect(new TextBlockCollector()).getSearchText().toLowerCase();
+        return strings.stream().map(String::toLowerCase).allMatch(rowText::contains);
+    }
+
+
+    /**
+     * Streams all entities which appear in a row where at least one cell has the provided header and the provided value.
+     *
+     * @param header the header value to search for
+     * @param value  the string which the table cell should contain
+     * @return a stream of all entities, which appear in a row where at least one cell has the provided header and the provided value.
+     */
+    public Stream<RedactionEntity> streamEntitiesWhereRowHasHeaderAndValue(String header, String value) {
+
+        List<Integer> vertebrateStudyCols = streamHeaders().filter(headerNode -> headerNode.containsString(header)).map(TableCell::getCol).toList();
+        return streamTableCells().filter(tableCellNode -> vertebrateStudyCols.stream()
+                .anyMatch(vertebrateStudyCol -> getCell(tableCellNode.getRow(), vertebrateStudyCol).containsString(value))).map(TableCell::getEntities).flatMap(Collection::stream);
+    }
+
+
+    /**
+     * Streams all entities which appear in a row where at least one cell has the provided header and any provided value.
+     *
+     * @param header the header value to search for
+     * @param values the strings which the table cell should contain
+     * @return a stream of all entities, which appear in a row where at least one cell has the provided header and any provided value.
+     */
+    public Stream<RedactionEntity> streamEntitiesWhereRowHasHeaderAndAnyValue(String header, List<String> values) {
+
+        List<Integer> vertebrateStudyCols = streamHeaders().filter(headerNode -> headerNode.containsString(header)).map(TableCell::getCol).toList();
+        return streamTableCells().filter(tableCellNode -> vertebrateStudyCols.stream()
+                        .anyMatch(vertebrateStudyCol -> getCell(tableCellNode.getRow(), vertebrateStudyCol).containsAnyString(values)))
+                .map(TableCell::getEntities)
+                .flatMap(Collection::stream);
+    }
+
+
+    /**
+     * Streams all entities in this table, that appear in a row, which contains at least one entity with any of the provided types.
+     *
+     * @param types type strings to check whether a row contains an entity like them
+     * @return Stream of all entities in this table, that appear in a row, which contains at least one entity with any of the provided types.
+     */
+    public Stream<RedactionEntity> streamEntitiesWhereRowContainsEntitiesOfType(List<String> types) {
+
+        List<Integer> rowsWithEntityOfType = getEntities().stream()
+                .filter(redactionEntity -> types.stream().anyMatch(type -> type.equals(redactionEntity.getType())))
+                .map(RedactionEntity::getIntersectingNodes)
+                .filter(node -> node instanceof TableCell)
+                .map(node -> (TableCell) node)
+                .map(TableCell::getRow)
+                .toList();
+
+        return rowsWithEntityOfType.stream().flatMap(this::streamRow).map(TableCell::getEntities).flatMap(Collection::stream);
+    }
+
+
+    /**
+     * Returns a TableCell at the provided row and column location.
+     *
+     * @param row int representing the row, must be smaller than numberOfRows
+     * @param col int representing the col, must be smaller than numberOfCols
+     * @return TableCell at the provided location in the table
+     */
+    public TableCell getCell(int row, int col) {
+
+        if (numberOfRows - row < 0 || numberOfCols - col < 0) {
+            throw new IllegalArgumentException(format("row %d, col %d is out of bounds for number of rows of %d and number of cols %d", row, col, numberOfRows, numberOfCols));
+        }
+        int idx = row * numberOfCols + col;
+        return (TableCell) documentTree.getEntryById(treeId).getChildren().get(idx).getNode();
+    }
+
+
+    /**
+     * Streams all TableCells in this Table row-wise.
+     *
+     * @return Stream of all TableCells
+     */
+    public Stream<TableCell> streamTableCells() {
+
+        return streamChildren().map(node -> (TableCell) node);
+    }
+
+
+    /**
+     * Streams all TableCells in this Table which have the provided header row-wise.
+     *
+     * @return Stream of all TableCells which have the provided header
+     */
+    public Stream<TableCell> streamTableCellsWithHeader(String header) {
+
+        return streamHeaders().filter(tableCellNode -> tableCellNode.buildTextBlock().getSearchText().contains(header))
+                .map(TableCell::getCol)
+                .flatMap(this::streamCol)
+                .filter(tableCellNode -> !tableCellNode.isHeader());
+    }
+
+
+    /**
+     * Streams all TableCells belonging to the provided column from top down.
+     *
+     * @param col int representing the column
+     * @return Stream of all TableCell in the provided column
+     */
+    public Stream<TableCell> streamCol(int col) {
+
+        return IntStream.range(0, numberOfRows).boxed().map(row -> getCell(row, col));
+    }
+
+
+    /**
+     * Streams all TableCells belonging to the provided row from left to right.
+     *
+     * @param row int representing the row
+     * @return Stream of all TableCell in the provided row
+     */
+    public Stream<TableCell> streamRow(int row) {
+
+        return IntStream.range(0, numberOfCols).boxed().map(col -> getCell(row, col));
+    }
+
+
+    /**
+     * Streams all TableCells row-wise and filters them with header == true.
+     *
+     * @return Stream of all TableCells with header == true
+     */
+    public Stream<TableCell> streamHeaders() {
+
+        return streamTableCells().filter(TableCell::isHeader);
+    }
+
+
+    /**
+     * Streams all TableCells of the provided row and column and filters them with header == true.
+     *
+     * @param row int representing the row
+     * @param col int representing the column
+     * @return Stream of all TableCells with header == true in the provided row or col
+     */
+    public Stream<TableCell> streamHeadersForCell(int row, int col) {
+
+        return Stream.concat(streamRow(row), streamCol(col)).filter(TableCell::isHeader);
+    }
+
+
+    /**
+     * Streams all Headers and checks if any equal the provided string.
+     *
+     * @param header string to check the headers for
+     * @return true, if at least one header equals the provided string
+     */
+    public boolean hasHeader(String header) {
+
+        return streamHeaders().anyMatch(tableCellNode -> tableCellNode.buildTextBlock().getSearchText().strip().equals(header));
+    }
+
+
+    /**
+     * Checks if this table has a column with the provided header and any of the table cells in that column contain the provided value.
+     *
+     * @param header string to find header cells
+     * @param value  string to check cells with provided header
+     * @return true, if this table has a column with the provided header and any of the table cells in that column contain the provided value
+     */
+    public boolean hasRowWithHeaderAndValue(String header, String value) {
+
+        return streamTableCellsWithHeader(header).anyMatch(tableCellNode -> tableCellNode.containsString(value));
+    }
+
+
+    /**
+     * Checks if this table has a column with the provided header and any of the table cells in that column contains any of the provided values.
+     *
+     * @param header string to find header cells
+     * @param values List of strings to check cells with provided header
+     * @return true, if this table has a column with the provided header and any of the table cells in that column contains any of the provided values.
+     */
+    public boolean hasRowWithHeaderAndAnyValue(String header, List<String> values) {
+
+        return streamTableCellsWithHeader(header).anyMatch(tableCellNode -> tableCellNode.containsAnyString(values));
+    }
+
+
+    /**
+     * Finds all entities of the provided type, which appear in the same row that the provided entity appears in.
+     *
+     * @param type            the type of entities to search for
+     * @param redactionEntity the entity, which appears in the row to search
+     * @return List of all entities of the provided type, which appear in the same row that the provided entity appears in.
+     */
+    public List<RedactionEntity> getEntitiesOfTypeInSameRow(String type, RedactionEntity redactionEntity) {
+
+        return redactionEntity.getIntersectingNodes()
+                .stream()
+                .filter(node -> node instanceof TableCell)
+                .map(node -> (TableCell) node)
+                .flatMap(tableCellNode -> streamRow(tableCellNode.getRow()))
+                .map(cell -> cell.getEntitiesOfType(type))
+                .flatMap(Collection::stream)
+                .toList();
+    }
+
+
+    @Override
+    public TextBlock buildTextBlock() {
+
+        if (textBlock == null) {
+            textBlock = streamAllSubNodes().filter(SemanticNode::isLeaf).map(SemanticNode::getLeafTextBlock).collect(new TextBlockCollector());
+        }
+        return textBlock;
+    }
+
+
+    @Override
+    public String toString() {
+
+        return treeId.toString() + ": " + NodeType.TABLE + ": #cols: " + numberOfCols + ", #rows: " + numberOfRows + ", " + buildTextBlock().buildSummary();
+    }
+
+}
--- a/Show More
+++ b/Show More
Author	SHA1	Message	Date
Kilian Schuettler	643e2e81d4	RED-6009 - Document Tree Structure * added gaps in Redaction Positions, if they are farther than five character widths apart	2023-05-10 16:45:23 +02:00
Kilian Schuettler	ce0a65c06a	RED-6009 - Document Tree Structure * refactored Visualization Class	2023-05-10 10:41:59 +02:00
Kilian Schuettler	f0144799c5	RED-6009 - Document Tree Structure * fixed rules to include vertebrate study flags	2023-05-09 17:20:33 +02:00
Kilian Schuettler	0443e21858	RED-6009 - Document Tree Structure * added more javadoc * fixed build	2023-05-09 16:50:52 +02:00
Kilian Schuettler	c151b4a94b	RED-6009 - Document Tree Structure * renamed terminal to leaf	2023-05-09 14:32:53 +02:00
Kilian Schuettler	e31a7b1a2f	RED-6009 - Document Tree Structure * added more javadoc	2023-05-09 14:12:12 +02:00
Kilian Schuettler	088a5cde1a	RED-6009 - Document Tree Structure * fixed javadoc and delombok plugins * removed unused classes * refactored names	2023-05-09 11:02:31 +02:00
Kilian Schuettler	c68c6f9ba9	RED-6009 - Document Tree Structure * updated rules to production syngenta rules	2023-05-05 14:40:37 +02:00
Kilian Schuettler	bdd0692df9	RED-6009 - Document Tree Structure * implemented table merging when no headers present	2023-05-05 14:31:48 +02:00
Kilian Schuettler	06df8bd47f	RED-6009 - Document Tree Structure * removed ner entity validation due to poor performance * filtered ner entities during reanalysis to improve performance	2023-05-05 14:31:48 +02:00
Kilian Schuettler	e193cc67d5	RED-6009 - Document Tree Structure * added logging for performance	2023-05-05 14:31:48 +02:00
Kilian Schuettler	826fce455f	RED-6009 - Document Tree Structure * moved NER Entity combination to drools	2023-05-05 14:31:48 +02:00
Kilian Schuettler	2f63627b60	RED-6009 - Document Tree Structure * improved logging for performance monitoring	2023-05-05 14:31:48 +02:00
Kilian Schuettler	1a6e678b00	RED-6009 - Document Tree Structure * remove Add ManualRedactions handling	2023-05-05 14:31:48 +02:00
Kilian Schuettler	cd471a1547	RED-6009 - Document Tree Structure * fix Ner Adapter	2023-05-05 14:31:48 +02:00
Kilian Schuettler	746f789be1	RED-6009 - Document Tree Structure * implement ManualRedactionEntry Handling	2023-05-05 14:31:48 +02:00
Kilian Schuettler	3f5e6dae66	RED-6009 - Document Tree Structure * fix parse imageType * fix getMatchedRule * remove trailing separator from Ai Entities	2023-05-05 14:31:48 +02:00
Kilian Schuettler	fdfb4c4071	RED-6009 - Document Tree Structure * fix rebased tests	2023-05-05 14:31:48 +02:00
Kilian Schuettler	cdab1d6945	RED-6009 - Document Tree Structure * fix nullpointer, if empty entity falls on end of TextBlock	2023-05-05 14:31:48 +02:00
Kilian Schuettler	0d5ef87d3e	RED-6009 - Document Tree Structure * image Type to lower case	2023-05-05 14:31:48 +02:00
Kilian Schuettler	37a3ee8d54	RED-6009 - Document Tree Structure * fix NerEntities CBI_address combination	2023-05-05 14:31:48 +02:00
Kilian Schuettler	f891e02935	RED-6009 - Document Tree Structure * minor rule fix	2023-05-05 14:31:48 +02:00
Kilian Schuettler	6d4326fb04	RED-6009 - Document Tree Structure * fix build	2023-05-05 14:31:48 +02:00
Kilian Schuettler	06163e1003	RED-6009: Document Tree Structure * fixed build * slight refactor of AnnotationService	2023-05-05 14:31:46 +02:00
Kilian Schuettler	5c4c8218f5	RED-6009: Document Tree Structure * added a test for removing entities by ManualResizeRedaction in ReAnalysis * Fixed Set logic when removing and readding a resized entity * Refactored RedactionLog change computation	2023-05-05 14:31:01 +02:00
Kilian Schuettler	22de0605d2	RED-6009: Document Tree Structure * wired ManualRedactionApplicationService to Drools Container	2023-05-05 14:31:01 +02:00
Kilian Schuettler	3ce08a34a6	RED-6009: Document Tree Structure * changed redactionLog Rectangle mapping to match UI	2023-05-05 14:31:01 +02:00
Kilian Schuettler	2f9b023bb4	RED-6009: Document Tree Structure * mapped pages back into ImageNodes in the graphMapper * fixed path in the noExceptionShouldBeThrownForAnyFilesTest	2023-05-05 14:31:01 +02:00
Kilian Schuettler	c7b08410e1	RED-6009: Document Tree Structure * optimized table access functions	2023-05-05 14:31:01 +02:00
Kilian Schuettler	57799c5f46	RED-6009: Document Tree Structure * pmd fixes	2023-05-05 14:31:01 +02:00
Kilian Schuettler	8b7365c947	RED-6009: Document Tree Structure * added a TableFactory to refactor DocumentGraphFactory * updated all enabled tests to run with new structure * moved PdfTableCells to classification package	2023-05-05 14:31:01 +02:00
Kilian Schuettler	c0aac985c1	RED-6009: Document Tree Structure renamed old TextBlock and Table Classes added subsections to enable separate rules for tables and text-only sections added ManualRedaction Tests finished default rules	2023-05-05 14:31:01 +02:00
Kilian Schuettler	1c0f7e308b	RED-6009: Document Tree Structure * refactored package structure	2023-05-05 14:31:01 +02:00
Kilian Schuettler	9c48d36b05	RED-6369: Rules Refactor * minor renaming * added NerEntityAdapter to include the Ner Entities * moved combine logic outside of rules	2023-05-05 14:31:01 +02:00
Kilian Schuettler	8bbaa4c140	RED-6369: Rules Refactor * added id to ImageNode * added Manual Redaction Handling to Rules * minor rename	2023-05-05 14:31:01 +02:00
Kilian Schuettler	3e7081a87a	RED-6369: Rules Refactor * fix pmd errors	2023-05-05 14:31:01 +02:00
Kilian Schuettler	c009790336	RED-6369: Rules Refactor * fixed builkd by disabling failing tests	2023-05-05 14:31:01 +02:00
Kilian Schuettler	09525f9b35	RED-6369: Rules Refactor *removed manualredaction handling for further analysis of how it actually works	2023-05-05 14:31:01 +02:00
Kilian Schuettler	282ee7652d	RED-6369: Rules Refactor * marked entities as resized if a resizeRedaction with same id exists	2023-05-05 14:31:01 +02:00
Kilian Schuettler	aa86e94f70	RED-6369: Rules Refactor * refactored documentGraph.tableOfContents to include document as root	2023-05-05 14:31:01 +02:00
Kilian Schuettler	9d8e544d0c	RED-6369: Rules Refactor * fixed AnnotationService with new Coordinates * fixed bug in split boundary function	2023-05-05 14:31:01 +02:00
Kilian Schuettler	59827eab81	RED-6369: Rules Refactor * refactored DocumentData to 4 different jsons to make reading only text possible * wip integration of new rules into workflow	2023-05-05 14:31:01 +02:00
Kilian Schuettler	ea0181d50f	RED-6369: Rules Refactor *added betweenStrings to entityCreationService	2023-05-05 14:31:01 +02:00
Kilian Schuettler	6402151311	RED-6369: Rules Refactor some refactors some comment cleanup	2023-05-05 14:31:01 +02:00
Kilian Schuettler	d804f0d0a2	RED-6369: Rules Refactor *fixed DocumentData Mapping	2023-05-05 14:31:01 +02:00
Kilian Schuettler	b6307759fb	RED-6369: Rules Refactor added some new Rules and Methods to support them integrated local dictionary searches into ruleflow using agenda-groups Generalized methods to be able to execute on any SemanticNode made Documentgraph implement SemanticNode	2023-05-05 14:31:01 +02:00
Kilian Schuettler	253e544724	RED-6368: Include Tables and Images into Prototype Document Structure *started Implementation of new rules	2023-05-05 14:31:01 +02:00
Kilian Schuettler	d04a576397	RED-6368: Include Tables and Images into Prototype Document Structure rebased to master and included tenant capabilities added separator validation to entity search *added EntityPosition class to bridge the gap to the redactionLogEntry	2023-05-05 14:31:01 +02:00
Kilian Schuettler	33d47e5202	RED-6368: Include Tables and Images into Prototype Document Structure *added some starting rules	2023-05-05 14:31:01 +02:00
Kilian Schuettler	dc76e581c4	RED-6368: Include Tables and Images into Prototype Document Structure added Tables And TableCell added BoundingBox calculation function *added images to structure without restructuring them, some images are still added twice	2023-05-05 14:31:01 +02:00
Kilian Schuettler	f3bfe95ea7	RED-6093: Prototype document structure refactored File Structure started refactor of original rules	2023-05-05 14:31:01 +02:00
Kilian Schuettler	53fa4f2851	RED-6093: Prototype document structure refactored Nodes added some tests *wip	2023-05-05 14:31:01 +02:00
deiflaender	64c54e73fb	RED-6093: Prototype document structurewip dependency upgrade	2023-05-05 14:31:01 +02:00
Kilian Schuettler	7c024a579c	RED-6093: Prototype document structure wip	2023-05-05 14:31:01 +02:00
Kilian Schuettler	3aa2f84f08	RED-6093: Prototype find entities in rules *added improved string to text position mapping	2023-05-05 14:31:01 +02:00
Kilian Schuettler	18dc68714b	RED-6093: Prototype find entities in rules *added a prototype paragraph rule	2023-05-05 14:31:01 +02:00
deiflaender	c4017e8e5e	RED-6093: Prototype find entities in rules	2023-05-05 14:31:01 +02:00