Changes

* added linebreaks to ParagraphData
* moved List<String> cellText to List<ParagraphData> cellTexts

.gitlab-ci.yml

@@ -1,5 +1,3 @@
-variables:
-  SONAR_PROJECT_KEY: 'fforesight_layout-parser_AYd5quv2mRkBOCG22hvF'
 include:
   - project: 'gitlab/gitlab'
     ref: 'main'

layoutparser-service-image/pom.xml

@@ -4,23 +4,23 @@
     <modelVersion>4.0.0</modelVersion>
 
 
-
     <parent>
-        <groupId>com.iqser.red</groupId>
+        <groupId>com.knecon.fforesight</groupId>
         <artifactId>platform-docker-dependency</artifactId>
-        <version>1.2.0</version>
+        <version>0.1.0</version>
         <relativePath/>
     </parent>
 
     <groupId>com.knecon.fforesight</groupId>
     <artifactId>layoutparser-service-image</artifactId>
-    <version>1.0.0</version>
+    <version>0.1-SNAPSHOT</version>
     <packaging>pom</packaging>
 
     <properties>
         <service.server>layoutparser-service-server</service.server>
         <platform.jar>${service.server}.jar</platform.jar>
         <docker.skip.push>false</docker.skip.push>
+        <docker.image.prefix>ff</docker.image.prefix>
         <docker.image.name>${docker.image.prefix}/${service.server}</docker.image.name>
     </properties>
 

layoutparser-service/layoutparser-service-internal-api/pom.xml

@@ -8,21 +8,16 @@
     <parent>
         <groupId>com.knecon.fforesight</groupId>
         <artifactId>layoutparser-service</artifactId>
-        <version>1.0.0</version>
+        <version>0.1-SNAPSHOT</version>
     </parent>
 
     <artifactId>layoutparser-service-internal-api</artifactId>
 
     <dependencies>
-        <dependency>
-            <groupId>org.projectlombok</groupId>
-            <artifactId>lombok</artifactId>
-            <version>1.18.26</version>
-        </dependency>
         <dependency>
             <groupId>com.google.guava</groupId>
             <artifactId>guava</artifactId>
-            <version>31.1-jre</version>
+            <version>${guava.version}</version>
         </dependency>
     </dependencies>
 

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/redaction/AtomicPositionBlockData.java

@@ -1,4 +1,4 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.data;
+package com.knecon.fforesight.service.layoutparser.internal.api.data.redaction;
 
 import lombok.AccessLevel;
 import lombok.AllArgsConstructor;

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/redaction/AtomicTextBlockData.java

@@ -1,4 +1,6 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.data;
+package com.knecon.fforesight.service.layoutparser.internal.api.data.redaction;
+
+
 
 import lombok.AccessLevel;
 import lombok.AllArgsConstructor;
@@ -20,4 +22,6 @@ public class AtomicTextBlockData {
     int end;
     int[] lineBreaks;
 
+
+
 }

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/redaction/DocumentData.java

@@ -1,4 +1,4 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.data;
+package com.knecon.fforesight.service.layoutparser.internal.api.data.redaction;
 
 import lombok.AccessLevel;
 import lombok.AllArgsConstructor;
@@ -15,6 +15,7 @@ public class DocumentData {
     PageData[] pages;
     AtomicTextBlockData[] atomicTextBlocks;
     AtomicPositionBlockData[] atomicPositionBlocks;
-    TableOfContentsData tableOfContents;
+    DocumentTreeData documentTreeData;
+
 
 }

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/redaction/DocumentTreeData.java

@@ -1,11 +1,9 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.data;
+package com.knecon.fforesight.service.layoutparser.internal.api.data.redaction;
 
 import java.util.List;
 import java.util.Map;
 import java.util.stream.Stream;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.NodeType;
-
 import lombok.AccessLevel;
 import lombok.AllArgsConstructor;
 import lombok.Builder;
@@ -19,7 +17,7 @@ import lombok.experimental.FieldDefaults;
 @AllArgsConstructor
 @NoArgsConstructor
 @FieldDefaults(level = AccessLevel.PRIVATE)
-public class TableOfContentsData {
+public class DocumentTreeData {
 
     EntryData root;
 
@@ -29,9 +27,9 @@ public class TableOfContentsData {
         if (tocId.isEmpty()) {
             return root;
         }
-        EntryData entry = root.subEntries.get(tocId.get(0));
+        EntryData entry = root.children.get(tocId.get(0));
         for (int id : tocId.subList(1, tocId.size())) {
-            entry = entry.subEntries.get(id);
+            entry = entry.children.get(id);
         }
         return entry;
     }
@@ -39,7 +37,7 @@ public class TableOfContentsData {
 
     public Stream<EntryData> streamAllEntries() {
 
-        return Stream.concat(Stream.of(root), root.subEntries.stream()).flatMap(TableOfContentsData::flatten);
+        return Stream.concat(Stream.of(root), root.children.stream()).flatMap(DocumentTreeData::flatten);
     }
 
 
@@ -51,7 +49,7 @@ public class TableOfContentsData {
 
     private static Stream<EntryData> flatten(EntryData entry) {
 
-        return Stream.concat(Stream.of(entry), entry.subEntries.stream().flatMap(TableOfContentsData::flatten));
+        return Stream.concat(Stream.of(entry), entry.children.stream().flatMap(DocumentTreeData::flatten));
     }
 
 
@@ -62,19 +60,18 @@ public class TableOfContentsData {
     public static class EntryData {
 
         NodeType type;
-        int[] tocId;
-        Long[] atomicBlocks;
-        Long[] pages;
+        int[] treeId;
+        Long[] atomicBlockIds;
+        Long[] pageNumbers;
         Map<String, String> properties;
-        List<EntryData> subEntries;
-
+        List<EntryData> children;
 
         @Override
         public String toString() {
 
             StringBuilder sb = new StringBuilder();
             sb.append("[");
-            for (int i : tocId) {
+            for (int i : treeId) {
                 sb.append(i);
                 sb.append(",");
             }
@@ -83,7 +80,7 @@ public class TableOfContentsData {
 
             sb.append(type);
             sb.append(" atbs = ");
-            sb.append(atomicBlocks.length);
+            sb.append(atomicBlockIds.length);
 
             return sb.toString();
         }

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/redaction/NodeType.java

@@ -1,4 +1,6 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes;
+package com.knecon.fforesight.service.layoutparser.internal.api.data.redaction;
+
+import java.util.Locale;
 
 public enum NodeType {
     DOCUMENT,
@@ -9,5 +11,11 @@ public enum NodeType {
     TABLE_CELL,
     IMAGE,
     HEADER,
-    FOOTER
+    FOOTER;
+
+
+    public String toString() {
+
+        return this.name().charAt(0) + this.name().substring(1).toLowerCase(Locale.ROOT);
+    }
 }

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/redaction/PageData.java

@@ -1,4 +1,4 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.data;
+package com.knecon.fforesight.service.layoutparser.internal.api.data.redaction;
 
 import lombok.AccessLevel;
 import lombok.AllArgsConstructor;

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/taas/ParagraphData.java

@@ -0,0 +1,21 @@
+package com.knecon.fforesight.service.layoutparser.internal.api.data.taas;
+
+import java.util.List;
+
+import lombok.Builder;
+import lombok.Data;
+
+@Data
+@Builder
+public class ParagraphData {
+
+    private String text;
+    List<Range> boldTextBoundaries;
+    List<Range> italicTextBoundaries;
+    List<Integer> linebreaks;
+    private String classification;
+
+    private String orientation;
+    private int textDirection;
+
+}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/taas/Range.java

@@ -0,0 +1,5 @@
+package com.knecon.fforesight.service.layoutparser.internal.api.data.taas;
+
+public record Range(int start, int end) {
+
+}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/taas/ResearchDocumentData.java

@@ -0,0 +1,16 @@
+package com.knecon.fforesight.service.layoutparser.internal.api.data.taas;
+
+import java.util.List;
+
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.Data;
+
+@Builder
+@Data
+@AllArgsConstructor
+public class ResearchDocumentData {
+
+    String originalFile;
+    List<StructureObject> structureObjects;
+}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/taas/RowData.java

@@ -0,0 +1,15 @@
+package com.knecon.fforesight.service.layoutparser.internal.api.data.taas;
+
+import java.util.List;
+
+import lombok.AllArgsConstructor;
+import lombok.Data;
+
+@Data
+@AllArgsConstructor
+public class RowData {
+
+    boolean header;
+    List<ParagraphData> cellText;
+    float[] bBox;
+}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/taas/StructureObject.java

@@ -0,0 +1,19 @@
+package com.knecon.fforesight.service.layoutparser.internal.api.data.taas;
+
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.Data;
+
+@Data
+@Builder
+@AllArgsConstructor
+public class StructureObject {
+
+    Integer structureObjectNumber;
+    int page;
+    int stringOffset;
+    float[] boundingBox;
+    ParagraphData paragraph;
+    TableData table;
+
+}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/taas/TableData.java

@@ -0,0 +1,15 @@
+package com.knecon.fforesight.service.layoutparser.internal.api.data.taas;
+
+import java.util.List;
+
+import lombok.AllArgsConstructor;
+import lombok.Data;
+
+@Data
+@AllArgsConstructor
+public class TableData {
+
+    List<RowData> rowData;
+    Integer numberOfCols;
+    Integer numberOfRows;
+}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/graph/DocumentGraph.java

@@ -1,101 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.graph;
-
-import java.awt.geom.Rectangle2D;
-import java.util.Collections;
-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.knecon.fforesight.service.layoutparser.internal.api.graph.entity.EntityNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.NodeType;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.PageNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.SectionNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.SemanticNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.textblock.TextBlock;
-import com.knecon.fforesight.service.layoutparser.internal.api.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 DocumentGraph implements SemanticNode {
-
-    Set<PageNode> pages;
-    TableOfContents tableOfContents;
-    Integer numberOfPages;
-    TextBlock textBlock;
-
-
-    public TextBlock buildTextBlock() {
-
-        return streamTerminalTextBlocksInOrder().collect(new TextBlockCollector());
-    }
-
-
-    public List<SectionNode> getMainSections() {
-
-        return streamChildren().filter(node -> node instanceof SectionNode).map(node -> (SectionNode) node).collect(Collectors.toList());
-    }
-
-
-    public Stream<TextBlock> streamTerminalTextBlocksInOrder() {
-
-        return streamAllNodes().filter(SemanticNode::isTerminal).map(SemanticNode::getTerminalTextBlock);
-    }
-
-
-    public Set<EntityNode> getEntities() {
-
-        return streamAllSubNodes().map(SemanticNode::getEntities).flatMap(Set::stream).collect(Collectors.toUnmodifiableSet());
-    }
-
-
-    @Override
-    public List<Integer> getTocId() {
-
-        return Collections.emptyList();
-    }
-
-
-    @Override
-    public void setTocId(List<Integer> tocId) {
-
-        throw new UnsupportedOperationException("DocumentGraph is always the root of the Table of Contents");
-    }
-
-
-    private Stream<SemanticNode> streamAllNodes() {
-
-        return tableOfContents.streamAllEntriesInOrder().map(TableOfContents.Entry::getNode);
-    }
-
-
-    @Override
-    public String toString() {
-
-        return NodeType.DOCUMENT + ": " + buildTextBlock().buildSummary();
-    }
-
-
-    @Override
-    public Map<PageNode, Rectangle2D> getBBox() {
-
-        Map<PageNode, Rectangle2D> bBox = new HashMap<>();
-        for (PageNode page : pages) {
-            bBox.put(page, new Rectangle2D.Double(0, 0, page.getWidth(), page.getHeight()));
-        }
-        return bBox;
-    }
-
-}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/graph/TableOfContents.java

@@ -1,193 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.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.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.NodeType;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.SemanticNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.textblock.TextBlock;
-import com.knecon.fforesight.service.layoutparser.internal.api.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 TableOfContents {
-
-    private final Entry root;
-
-
-    public TableOfContents(DocumentGraph documentGraph) {
-
-        root = Entry.builder().tocId(Collections.emptyList()).type(NodeType.DOCUMENT).children(new LinkedList<>()).node(documentGraph).build();
-    }
-
-
-    public TextBlock buildTextBlock() {
-
-        return streamAllEntriesInOrder().map(Entry::getNode).filter(SemanticNode::isTerminal).map(SemanticNode::getTerminalTextBlock).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 entryExists(getParentId(tocId));
-    }
-
-
-    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(TableOfContents::flatten);
-    }
-
-
-    public Stream<Entry> streamAllSubEntriesInOrder(List<Integer> parentId) {
-
-        return getEntryById(parentId).getChildren().stream().flatMap(TableOfContents::flatten);
-    }
-
-
-    @Override
-    public String toString() {
-
-        return String.join("\n", streamAllEntriesInOrder().map(Entry::toString).toList());
-    }
-
-
-    public String toString(List<Integer> id) {
-
-        return String.join("\n", streamAllSubEntriesInOrder(id).map(Entry::toString).toList());
-    }
-
-
-    private static Stream<Entry> flatten(Entry entry) {
-
-        return Stream.concat(Stream.of(entry), entry.children.stream().flatMap(TableOfContents::flatten));
-    }
-
-
-    @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();
-        }
-
-    }
-
-}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/graph/entity/EntityNode.java

@@ -1,76 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.graph.entity;
-
-import java.util.Collections;
-import java.util.List;
-import java.util.Set;
-
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.Boundary;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.PageNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.SemanticNode;
-
-public interface EntityNode {
-
-    /**
-     * This represents the text, which is contained within the boundary of the Entity.
-     *
-     * @return String
-     */
-    String getValue();
-
-
-    /**
-     * The Boundary primarily defines the Entity, all other values may be inferred from it.
-     *
-     * @return Boundary, uniquely identifying this Entity
-     */
-    Boundary getBoundary();
-
-
-    /**
-     * The deepest fully containing node represents the node which is the deepest node in the document tree structure,
-     * whose boundary also fully contains the boundary of this entity.
-     *
-     * @return the deepest fully containing node
-     */
-    SemanticNode getDeepestFullyContainingNode();
-
-
-    /**
-     * The intersecting nodes represent all nodes, whose boundary intersects the boundary of this entity.
-     *
-     * @return all intersecting Nodes
-     */
-    List<SemanticNode> getIntersectingNodes();
-
-
-    void setDeepestFullyContainingNode(SemanticNode semanticNode);
-
-
-    void addIntersectingNode(SemanticNode semanticNode);
-
-
-    void setIntersectingNodes(List<SemanticNode> semanticNodes);
-
-
-    /**
-     * @return all pages this entity intersects.
-     */
-    Set<PageNode> getPages();
-
-
-    void setPages(Set<PageNode> pages);
-
-
-    /**
-     * removes all occurrences of this node in the graph and resets all graph specific fields.
-     */
-    default void removeFromGraph() {
-
-        getIntersectingNodes().forEach(node -> node.getEntities().remove(this));
-        getPages().forEach(page -> page.getEntities().remove(this));
-        setPages(Collections.emptySet());
-        setDeepestFullyContainingNode(null);
-        setIntersectingNodes(Collections.emptyList());
-    }
-
-}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/graph/entity/EntityPosition.java

@@ -1,45 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.graph.entity;
-
-import java.awt.geom.Rectangle2D;
-import java.nio.charset.StandardCharsets;
-import java.util.List;
-
-import com.google.common.hash.Hashing;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.PageNode;
-
-import lombok.AccessLevel;
-import lombok.Builder;
-import lombok.Data;
-import lombok.experimental.FieldDefaults;
-
-@Data
-@Builder
-@FieldDefaults(level = AccessLevel.PRIVATE)
-public class EntityPosition {
-
-    PageNode pageNode;
-    List<Rectangle2D> rectanglePerLine;
-
-
-    public String getId() {
-
-        return String.valueOf(hashCode());
-    }
-
-
-    @Override
-    public int hashCode() {
-
-        StringBuilder sb = new StringBuilder();
-        sb.append(pageNode.getNumber());
-        rectanglePerLine.forEach(r -> sb.append(r.getX()).append(r.getY()).append(r.getWidth()).append(r.getHeight()));
-        return Hashing.murmur3_128().hashString(sb.toString(), StandardCharsets.UTF_8).hashCode();
-    }
-
-    @Override
-    public boolean equals(Object o) {
-
-        return o instanceof EntityPosition && o.hashCode() == this.hashCode();
-    }
-
-}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/graph/nodes/FooterNode.java

@@ -1,53 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes;
-
-import java.util.HashSet;
-import java.util.List;
-import java.util.Set;
-
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.TableOfContents;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.entity.EntityNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.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 FooterNode implements SemanticNode {
-
-    List<Integer> tocId;
-    TextBlock terminalTextBlock;
-
-    @Builder.Default
-    boolean terminal = true;
-
-    @EqualsAndHashCode.Exclude
-    TableOfContents tableOfContents;
-
-    @Builder.Default
-    @EqualsAndHashCode.Exclude
-    Set<EntityNode> entities = new HashSet<>();
-
-
-    @Override
-    public TextBlock buildTextBlock() {
-
-        return terminalTextBlock;
-    }
-
-
-    @Override
-    public String toString() {
-
-        return tocId + ": " + NodeType.FOOTER + ": " + terminalTextBlock.buildSummary();
-    }
-
-}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/graph/nodes/HeaderNode.java

@@ -1,53 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes;
-
-import java.util.HashSet;
-import java.util.List;
-import java.util.Set;
-
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.TableOfContents;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.entity.EntityNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.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 HeaderNode implements SemanticNode {
-
-    List<Integer> tocId;
-    TextBlock terminalTextBlock;
-
-    @Builder.Default
-    boolean terminal = true;
-
-    @EqualsAndHashCode.Exclude
-    TableOfContents tableOfContents;
-
-    @Builder.Default
-    @EqualsAndHashCode.Exclude
-    Set<EntityNode> entities = new HashSet<>();
-
-
-    @Override
-    public TextBlock buildTextBlock() {
-
-        return terminalTextBlock;
-    }
-
-
-    @Override
-    public String toString() {
-
-        return tocId + ": " + NodeType.HEADER + ": " + terminalTextBlock.buildSummary();
-    }
-
-}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/graph/nodes/HeadlineNode.java

@@ -1,60 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes;
-
-import java.util.HashSet;
-import java.util.List;
-import java.util.Set;
-
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.TableOfContents;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.entity.EntityNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.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 HeadlineNode implements SemanticNode {
-
-    List<Integer> tocId;
-    TextBlock terminalTextBlock;
-
-    @Builder.Default
-    boolean terminal = true;
-
-    @EqualsAndHashCode.Exclude
-    TableOfContents tableOfContents;
-
-    @Builder.Default
-    @EqualsAndHashCode.Exclude
-    Set<EntityNode> entities = new HashSet<>();
-
-
-    @Override
-    public TextBlock buildTextBlock() {
-
-        return terminalTextBlock;
-    }
-
-
-    @Override
-    public String toString() {
-
-        return tocId + ": " + NodeType.HEADLINE + ": " + terminalTextBlock.buildSummary();
-    }
-
-
-    @Override
-    public SemanticNode getHeadline() {
-
-        return this;
-    }
-
-}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/graph/nodes/ImageNode.java

@@ -1,87 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.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.knecon.fforesight.service.layoutparser.internal.api.graph.TableOfContents;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.entity.EntityNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.textblock.TextBlock;
-import com.knecon.fforesight.service.layoutparser.internal.api.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 ImageNode implements SemanticNode {
-
-    List<Integer> tocId;
-
-    ImageType imageType;
-    boolean transparency;
-    Rectangle2D position;
-
-
-    boolean redaction;
-    boolean ignored;
-
-    @Builder.Default
-    String redactionReason = "";
-    @Builder.Default
-    String legalBasis = "";
-    @Builder.Default
-    int matchedRule = -1;
-
-    @EqualsAndHashCode.Exclude
-    PageNode page;
-
-    @EqualsAndHashCode.Exclude
-    TableOfContents tableOfContents;
-
-    @Builder.Default
-    @EqualsAndHashCode.Exclude
-    Set<EntityNode> entities = new HashSet<>();
-
-    @Override
-    public TextBlock buildTextBlock() {
-
-        return streamAllSubNodes().filter(SemanticNode::isTerminal).map(SemanticNode::getTerminalTextBlock).collect(new TextBlockCollector());
-    }
-
-
-    @Override
-    public Set<PageNode> getPages() {
-
-        return Collections.singleton(page);
-    }
-
-
-    @Override
-    public String toString() {
-
-        return tocId + ": " + NodeType.IMAGE + ": " + imageType.toString() + " " + position;
-    }
-
-
-    @Override
-    public Map<PageNode, Rectangle2D> getBBox() {
-
-        Map<PageNode, Rectangle2D> bBoxPerPage = new HashMap<>();
-        bBoxPerPage.put(page, position);
-        return bBoxPerPage;
-    }
-
-}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/graph/nodes/ImageType.java

@@ -1,9 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes;
-
-public enum ImageType {
-    LOGO,
-    FORMULA,
-    SIGNATURE,
-    OTHER,
-    OCR
-}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/graph/nodes/PageNode.java

@@ -1,71 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes;
-
-import java.util.HashSet;
-import java.util.List;
-import java.util.Set;
-
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.entity.EntityNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.textblock.TextBlock;
-import com.knecon.fforesight.service.layoutparser.internal.api.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 PageNode {
-
-    Integer number;
-    Integer height;
-    Integer width;
-    Integer rotation;
-
-    @EqualsAndHashCode.Exclude
-    List<SemanticNode> mainBody;
-    @EqualsAndHashCode.Exclude
-    HeaderNode header;
-    @EqualsAndHashCode.Exclude
-    FooterNode footer;
-
-    @Builder.Default
-    @EqualsAndHashCode.Exclude
-    Set<EntityNode> entities = new HashSet<>();
-
-    @Builder.Default
-    @EqualsAndHashCode.Exclude
-    Set<ImageNode> images = new HashSet<>();
-
-
-    public TextBlock getMainBodyTextBlock() {
-
-        return mainBody.stream().filter(SemanticNode::isTerminal).map(SemanticNode::getTerminalTextBlock).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 PageNode && o.hashCode() == this.hashCode();
-    }
-}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/graph/nodes/ParagraphNode.java

@@ -1,51 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes;
-
-import java.util.HashSet;
-import java.util.List;
-import java.util.Set;
-
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.TableOfContents;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.entity.EntityNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.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 ParagraphNode implements SemanticNode {
-
-    List<Integer> tocId;
-    TextBlock terminalTextBlock;
-
-    @Builder.Default
-    boolean terminal = true;
-
-    @EqualsAndHashCode.Exclude
-    TableOfContents tableOfContents;
-
-    @Builder.Default
-    @EqualsAndHashCode.Exclude
-    Set<EntityNode> entities = new HashSet<>();
-
-
-    @Override
-    public TextBlock buildTextBlock() {
-
-        return terminalTextBlock;
-    }
-
-
-    @Override
-    public String toString() {
-
-        return tocId + ": " + NodeType.PARAGRAPH + ": " + terminalTextBlock.buildSummary();
-    }
-
-}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/graph/nodes/SectionNode.java

@@ -1,63 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes;
-
-import java.util.HashSet;
-import java.util.List;
-import java.util.NoSuchElementException;
-import java.util.Set;
-
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.TableOfContents;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.entity.EntityNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.textblock.TextBlock;
-import com.knecon.fforesight.service.layoutparser.internal.api.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 SectionNode implements SemanticNode {
-
-    List<Integer> tocId;
-
-    TextBlock textBlock;
-    @EqualsAndHashCode.Exclude
-    TableOfContents tableOfContents;
-
-    @Builder.Default
-    @EqualsAndHashCode.Exclude
-    Set<EntityNode> entities = new HashSet<>();
-
-
-    @Override
-    public TextBlock buildTextBlock() {
-
-        if (textBlock == null) {
-            textBlock = streamAllSubNodes().filter(SemanticNode::isTerminal).map(SemanticNode::getTerminalTextBlock).collect(new TextBlockCollector());
-        }
-        return textBlock;
-    }
-
-
-    @Override
-    public String toString() {
-
-        return tocId.toString() + ": " + NodeType.SECTION + ": " + buildTextBlock().buildSummary();
-    }
-
-
-    public HeadlineNode getHeadline() {
-
-        return streamChildren().filter(node -> node instanceof HeadlineNode)
-                .map(node -> (HeadlineNode) node)
-                .findFirst()
-                .orElseThrow(() -> new NoSuchElementException("ClassificationSection has no Headline!"));
-    }
-
-}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/graph/nodes/SemanticNode.java

@@ -1,275 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes;
-
-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.knecon.fforesight.service.layoutparser.internal.api.graph.Boundary;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.TableOfContents;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.entity.EntityNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.textblock.AtomicTextBlock;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.textblock.TextBlock;
-import com.knecon.fforesight.service.layoutparser.internal.api.services.RectangleTransformations;
-
-public interface SemanticNode {
-
-    /**
-     * Searches all Nodes located underneath this Node in the TableOfContents and concatenates their AtomicTextBlocks into a single TextBlockEntity.
-     * So, for a ClassificationSection all TextBlocks of Subsections, Paragraphs, and Tables are concatenated into a single TextBlockEntity
-     * If the Node is Terminal, the TerminalTextBlock will be returned instead.
-     *
-     * @return ClassificationTextBlock 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<EntityNode> 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 ClassificationTextBlock.
-     *
-     * @return Set of PageNodes this node appears on.
-     */
-    default Set<PageNode> getPages() {
-
-        return buildTextBlock().getPages();
-    }
-
-
-    /**
-     * @return the TableOfContents of the ClassificationDocument this node belongs to
-     */
-    TableOfContents getTableOfContents();
-
-
-    /**
-     * The id is a List of Integers uniquely identifying this node in the TableOfContents.
-     *
-     * @return the TableOfContents ID
-     */
-    List<Integer> getTocId();
-
-
-    /**
-     * This should only be used during graph construction.
-     *
-     * @param tocId List of Integers
-     */
-    void setTocId(List<Integer> tocId);
-
-
-    /**
-     * Traverses the Tree up, until it hits a HeadlineNode or hits a SectionNode which will then return the first HeadlineNode from its children.
-     * Throws NotFoundException if no Headline is found this way
-     *
-     * @return First HeadlineNode found
-     */
-    default SemanticNode getHeadline() {
-
-        return getParent().getHeadline();
-    }
-
-
-    /**
-     * @return boolean indicating wether this Node has a Parent in the TableOfContents
-     */
-    default boolean hasParent() {
-
-        return getTableOfContents().hasParentById(getTocId());
-    }
-
-
-    /**
-     * @return The SemanticNode representing the Parent in the TableOfContents
-     * throws NotFoundException, when no parent is present
-     */
-    default SemanticNode getParent() {
-
-        return getTableOfContents().getParentEntryById(getTocId()).getNode();
-    }
-
-
-    /**
-     * Terminal means a SemanticNode has direct access to a ClassificationTextBlock, by default this is false and must be overridden.
-     * Currently only Sections, Images, and Tables are not terminal.
-     * A TableCell might be Terminal depending on its area compared to the page.
-     *
-     * @return boolean, indicating if a Node has direct access to a ClassificationTextBlock
-     */
-    default boolean isTerminal() {
-
-        return false;
-    }
-
-
-    /**
-     * Terminal means a SemanticNode has direct access to a ClassificationTextBlock, by default this is false and must be overridden.
-     * Currently only Sections and Tables are not terminal.
-     *
-     * @return AtomicTextBlock
-     */
-    default TextBlock getTerminalTextBlock() {
-
-        throw new UnsupportedOperationException("Only terminal Nodes have access to TerminalTextBlocks!");
-    }
-
-
-    default void setTerminalTextBlock(TextBlock textBlock) {
-
-        throw new UnsupportedOperationException();
-    }
-
-
-    /**
-     * 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;
-        }
-    }
-
-
-    /**
-     * @return true, if this node's ClassificationTextBlock is not empty
-     */
-    default boolean hasText() {
-
-        return buildTextBlock().length() > 0;
-    }
-
-
-    /**
-     * @param string A String which the ClassificationTextBlock might contain
-     * @return true, if this node's ClassificationTextBlock contains the string
-     */
-    default boolean containsString(String string) {
-
-        return buildTextBlock().getSearchText().contains(string);
-    }
-
-
-    /**
-     * @param strings A List of Strings which the ClassificationTextBlock might contain
-     * @return true, if this node's ClassificationTextBlock contains any of the strings
-     */
-    default boolean containsAnyString(List<String> strings) {
-
-        return strings.stream().anyMatch(this::containsString);
-    }
-
-
-    /**
-     * This function is used during insertion of EntityNodes into the graph, it checks if the boundary of the EntityNode intersects or even contains the EntityNode.
-     * It sets the fields accordingly and recursively calls this function on all its children.
-     *
-     * @param entityNode EntityNode, which is being inserted into the graph
-     */
-    default void addThisToEntityIfIntersects(EntityNode entityNode) {
-
-        TextBlock textBlock = buildTextBlock();
-        if (textBlock.getBoundary().intersects(entityNode.getBoundary())) {
-
-            if (textBlock.containsBoundary(entityNode.getBoundary())) {
-                entityNode.setDeepestFullyContainingNode(this);
-            }
-
-            entityNode.addIntersectingNode(this);
-            streamChildren().forEach(node -> node.addThisToEntityIfIntersects(entityNode));
-        }
-    }
-
-
-    /**
-     * Streams all children located directly underneath this node in the TableOfContents.
-     *
-     * @return Stream of all children
-     */
-    default Stream<SemanticNode> streamChildren() {
-
-        return getTableOfContents().streamChildrenNodes(getTocId());
-    }
-
-
-    /**
-     * recursively streams all SemanticNodes located underneath this node in the TableOfContents in order.
-     *
-     * @return Stream of all SubNodes
-     */
-    default Stream<SemanticNode> streamAllSubNodes() {
-
-        return getTableOfContents().streamAllSubEntriesInOrder(getTocId()).map(TableOfContents.Entry::getNode);
-    }
-
-
-    /**
-     * @return Boundary of this Node's ClassificationTextBlock
-     */
-    default Boundary getBoundary() {
-
-        return buildTextBlock().getBoundary();
-    }
-
-
-    /**
-     * If this Node is Terminal it will calculate the boundingBox of its TerminalTextBlock, otherwise it will calculate the Union of the BoundingBoxes of all its Children.
-     * If called on the ClassificationDocument, it will return the cropbox of each page
-     *
-     * @return Rectangle2D fully encapsulating this Node for each page.
-     */
-    default Map<PageNode, Rectangle2D> getBBox() {
-
-        Map<PageNode, Rectangle2D> bBoxPerPage = new HashMap<>();
-        if (isTerminal()) {
-            return getBBoxFromTerminalTextBlock(bBoxPerPage);
-        }
-
-        return getBBoxFromChildren(bBoxPerPage);
-    }
-
-
-    /**
-     * TODO this does not yet work for sections spanning multiple columns.
-     *
-     * @param bBoxPerPage initial empty BoundingBox
-     * @return The union of the BoundingBoxes of all children
-     */
-    private Map<PageNode, Rectangle2D> getBBoxFromChildren(Map<PageNode, 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 ClassificationTextBlock of this node
-     */
-    private Map<PageNode, Rectangle2D> getBBoxFromTerminalTextBlock(Map<PageNode, Rectangle2D> bBoxPerPage) {
-
-        Map<PageNode, List<AtomicTextBlock>> atomicTextBlockPerPage = buildTextBlock().getAtomicTextBlocks().stream().collect(Collectors.groupingBy(AtomicTextBlock::getPage));
-        atomicTextBlockPerPage.forEach((page, atbs) -> bBoxPerPage.put(page, RectangleTransformations.bBoxUnionAtomicTextBlock(atbs)));
-        return bBoxPerPage;
-    }
-
-}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/graph/nodes/TableCellNode.java

@@ -1,92 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes;
-
-import java.awt.geom.Rectangle2D;
-import java.util.HashMap;
-import java.util.HashSet;
-import java.util.List;
-import java.util.Map;
-import java.util.Set;
-import java.util.stream.Stream;
-
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.TableOfContents;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.entity.EntityNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.textblock.TextBlock;
-import com.knecon.fforesight.service.layoutparser.internal.api.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 TableCellNode implements SemanticNode {
-
-    List<Integer> tocId;
-    int row;
-    int col;
-    boolean header;
-
-    Rectangle2D bBox;
-
-    @Builder.Default
-    boolean terminal = true;
-    TextBlock terminalTextBlock;
-
-    TextBlock textBlock;
-
-    @EqualsAndHashCode.Exclude
-    TableOfContents tableOfContents;
-
-    @Builder.Default
-    @EqualsAndHashCode.Exclude
-    Set<EntityNode> entities = new HashSet<>();
-
-
-    @Override
-    public Map<PageNode, Rectangle2D> getBBox() {
-
-        Map<PageNode, Rectangle2D> bBoxPerPage = new HashMap<>();
-        getPages().forEach(page -> bBoxPerPage.put(page, bBox));
-        return bBoxPerPage;
-    }
-
-
-    @Override
-    public TextBlock buildTextBlock() {
-
-        if (terminal) {
-            return terminalTextBlock;
-        }
-
-        if (textBlock == null) {
-            textBlock = streamAllSubNodes().filter(SemanticNode::isTerminal).map(SemanticNode::getTerminalTextBlock).collect(new TextBlockCollector());
-        }
-        return textBlock;
-    }
-
-
-    @Override
-    public String toString() {
-
-        return tocId + ": " + NodeType.TABLE_CELL + ": " + buildTextBlock().buildSummary();
-    }
-
-
-    public boolean hasHeader(String headerString) {
-
-        return getHeaders().anyMatch(header -> header.buildTextBlock().getSearchText().strip().equals(headerString));
-    }
-
-
-    private Stream<TableCellNode> getHeaders() {
-
-        TableNode tableNode = (TableNode) getParent();
-        return tableNode.streamHeadersForCell(row, col);
-    }
-
-}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/graph/nodes/TableNode.java

@@ -1,73 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes;
-
-import java.util.HashSet;
-import java.util.List;
-import java.util.Set;
-import java.util.stream.Stream;
-
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.TableOfContents;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.entity.EntityNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.textblock.TextBlock;
-import com.knecon.fforesight.service.layoutparser.internal.api.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 TableNode implements SemanticNode {
-
-    List<Integer> tocId;
-    TableOfContents tableOfContents;
-
-    Integer numberOfRows;
-    Integer numberOfCols;
-
-    TextBlock textBlock;
-
-    @Builder.Default
-    @EqualsAndHashCode.Exclude
-    Set<EntityNode> entities = new HashSet<>();
-
-
-    public Stream<TableCellNode> streamTableCells() {
-
-        return streamChildren().map(node -> (TableCellNode) node);
-    }
-
-
-    public Stream<TableCellNode> streamHeaders() {
-
-        return streamTableCells().filter(TableCellNode::isHeader);
-    }
-
-
-    public Stream<TableCellNode> streamHeadersForCell(int row, int col) {
-
-        return streamHeaders().filter(cell -> cell.getRow() == row || cell.getCol() == col);
-    }
-
-
-    @Override
-    public TextBlock buildTextBlock() {
-
-        if (textBlock == null) {
-            textBlock = streamAllSubNodes().filter(SemanticNode::isTerminal).map(SemanticNode::getTerminalTextBlock).collect(new TextBlockCollector());
-        }
-        return textBlock;
-    }
-
-
-    @Override
-    public String toString() {
-
-        return tocId.toString() + ": " + NodeType.TABLE + ": " + buildTextBlock().buildSummary();
-    }
-
-}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/graph/textblock/AtomicTextBlock.java

@@ -1,131 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.graph.textblock;
-
-import java.awt.geom.Rectangle2D;
-import java.util.List;
-
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.Boundary;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.entity.EntityPosition;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.PageNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.SemanticNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.services.RectangleTransformations;
-
-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 AtomicTextBlock implements TextBlock {
-
-    Long id;
-    Integer numberOnPage;
-    PageNode page;
-
-    //string coordinates
-    Boundary boundary;
-    String searchText;
-    List<Integer> lineBreaks;
-
-    //position coordinates
-    List<Integer> stringIdxToPositionIdx;
-    List<Rectangle2D> positions;
-
-    @EqualsAndHashCode.Exclude
-    SemanticNode parent;
-
-
-    @Override
-    public int numberOfLines() {
-
-        return lineBreaks.size() + 1;
-    }
-
-
-    public CharSequence getLine(int lineNumber) {
-
-        if (lineNumber >= numberOfLines() || lineNumber < 0) {
-            throw new IndexOutOfBoundsException(String.format("line %d out of range for AtomicTextBlock with %d lines", lineNumber, numberOfLines()));
-        }
-        if (lineNumber == 0) {
-            return subSequence(boundary.start(), lineBreaks.get(0) + boundary.start());
-        } else if (lineNumber == numberOfLines() - 1) {
-            return subSequence(lineBreaks.get(lineBreaks.size() - 1) + boundary.start(), boundary.end());
-        }
-        return subSequence(lineBreaks.get(lineNumber - 1) + boundary.start(), lineBreaks.get(lineNumber) + boundary.start());
-    }
-
-
-    @Override
-    public List<AtomicTextBlock> getAtomicTextBlocks() {
-
-        return List.of(this);
-    }
-
-
-    @Override
-    public int getNextLinebreak(int fromIndex) {
-
-        return lineBreaks.stream()//
-                .filter(linebreak -> linebreak > fromIndex - boundary.start()) //
-                .findFirst() //
-                .orElse(searchText.length()) + boundary.start();
-    }
-
-
-    @Override
-    public int getPreviousLinebreak(int fromIndex) {
-
-        return lineBreaks.stream()//
-                .filter(linebreak -> linebreak <= fromIndex - boundary.start())//
-                .reduce((a, b) -> b)//
-                .orElse(0) + boundary.start();
-    }
-
-
-    @Override
-    public Rectangle2D getPosition(int stringIdx) {
-
-        return positions.get(stringIdxToPositionIdx.get(stringIdx - boundary.start()));
-    }
-
-
-    @Override
-    public List<Rectangle2D> getPositions(Boundary stringBoundary) {
-
-        if (!containsBoundary(stringBoundary)) {
-            throw new IndexOutOfBoundsException(String.format("%s is out of bounds for %s", stringBoundary, this.boundary));
-        }
-
-        if (stringBoundary.end() == this.boundary.end()) {
-            return positions.subList(stringIdxToPositionIdx.get(stringBoundary.start() - this.boundary.start()), positions.size());
-        }
-
-        return positions.subList(stringIdxToPositionIdx.get(stringBoundary.start() - this.boundary.start()),
-                stringIdxToPositionIdx.get(stringBoundary.end() - this.boundary.start()));
-    }
-
-
-    public List<EntityPosition> getEntityPositionsPerPage(Boundary stringBoundary) {
-
-        List<Rectangle2D> positionsPerLine = stringBoundary.split(getLineBreaks().stream().map(lb -> lb + boundary.start()).filter(stringBoundary::contains).toList())
-                .stream()
-                .map(this::getPositions)
-                .map(RectangleTransformations::rectangleUnion)
-                .toList();
-
-        return List.of(EntityPosition.builder().rectanglePerLine(positionsPerLine).pageNode(page).build());
-    }
-
-
-    @Override
-    public String toString() {
-
-        return searchText;
-    }
-
-}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/mapper/DocumentGraphMapper.java

@@ -1,229 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.mapper;
-
-import static com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.NodeType.FOOTER;
-import static com.knecon.fforesight.service.layoutparser.internal.api.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.knecon.fforesight.service.layoutparser.internal.api.data.AtomicPositionBlockData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.AtomicTextBlockData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.DocumentData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.PageData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.TableOfContentsData;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.Boundary;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.DocumentGraph;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.TableOfContents;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.FooterNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.HeaderNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.HeadlineNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.ImageNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.PageNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.ParagraphNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.SectionNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.SemanticNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.TableCellNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.TableNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.textblock.AtomicTextBlock;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.textblock.TextBlock;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.textblock.TextBlockCollector;
-
-import lombok.experimental.UtilityClass;
-
-@UtilityClass
-public class DocumentGraphMapper {
-
-    public DocumentGraph toDocumentGraph(DocumentData documentData) {
-
-
-        DocumentGraph documentGraph = new DocumentGraph();
-        Context context = new Context(documentData,
-                new TableOfContents(documentGraph),
-                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.tableOfContents.getRoot().getChildren().addAll(buildEntries(documentData.getTableOfContents().getRoot().getSubEntries(), context));
-
-        documentGraph.setTableOfContents(context.tableOfContents);
-        documentGraph.setPages(new HashSet<>(context.pages));
-        documentGraph.setNumberOfPages(documentData.getPages().length);
-
-        documentGraph.setTextBlock(documentGraph.buildTextBlock());
-        return documentGraph;
-    }
-
-
-    private List<TableOfContents.Entry> buildEntries(List<TableOfContentsData.EntryData> entries,
-                                                     Context context) {
-
-        List<TableOfContents.Entry> newEntries = new LinkedList<>();
-        for (TableOfContentsData.EntryData entryData : entries) {
-
-            boolean terminal = isTerminal(entryData);
-            List<PageNode> pages = Arrays.stream(entryData.getPages()).map(pageNumber -> getPage(pageNumber, context)).toList();
-
-            SemanticNode node = switch (entryData.getType()) {
-                case SECTION -> buildSection(context);
-                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());
-                default -> throw new UnsupportedOperationException("Not yet implemented for type " + entryData.getType());
-            };
-
-            if (node.isTerminal()) {
-                TextBlock textBlock = toTextBlock(entryData.getAtomicBlocks(), context, node);
-                node.setTerminalTextBlock(textBlock);
-            }
-            List<Integer> tocId = Arrays.stream(entryData.getTocId()).boxed().toList();
-            node.setTocId(tocId);
-
-            if (entryData.getType() == HEADER) {
-                pages.forEach(page -> page.setHeader((HeaderNode) node));
-            } else if (entryData.getType() == FOOTER) {
-                pages.forEach(page -> page.setFooter((FooterNode) node));
-            } else {
-                pages.forEach(page -> page.getMainBody().add(node));
-            }
-            newEntries.add(TableOfContents.Entry.builder().tocId(tocId).type(entryData.getType()).children(buildEntries(entryData.getSubEntries(), context)).node(node).build());
-        }
-        return newEntries;
-    }
-
-
-    private HeadlineNode buildHeadline(Context context, boolean terminal) {
-
-        return HeadlineNode.builder().terminal(terminal).tableOfContents(context.tableOfContents()).build();
-    }
-
-
-    private static boolean isTerminal(TableOfContentsData.EntryData entryData) {
-
-        return entryData.getAtomicBlocks().length > 0;
-    }
-
-
-    private ImageNode buildImage(Context context, Map<String, String> properties) {
-
-        var builder = ImageNode.builder();
-        PropertiesMapper.parseImageProperties(properties, builder);
-        return builder.tableOfContents(context.tableOfContents()).build();
-    }
-
-
-    private TableCellNode buildTableCell(Context context, Map<String, String> properties, boolean terminal) {
-
-        TableCellNode.TableCellNodeBuilder builder = TableCellNode.builder();
-        PropertiesMapper.parseTableCellProperties(properties, builder);
-        return builder.terminal(terminal).tableOfContents(context.tableOfContents()).build();
-    }
-
-
-    private TableNode buildTable(Context context, Map<String, String> properties) {
-
-        TableNode.TableNodeBuilder builder = TableNode.builder();
-        PropertiesMapper.parseTableProperties(properties, builder);
-        return TableNode.builder().tableOfContents(context.tableOfContents()).build();
-    }
-
-
-    private FooterNode buildFooter(Context context, boolean terminal) {
-
-        return FooterNode.builder().terminal(terminal).tableOfContents(context.tableOfContents()).build();
-    }
-
-
-    private HeaderNode buildHeader(Context context, boolean terminal) {
-
-        return HeaderNode.builder().terminal(terminal).tableOfContents(context.tableOfContents()).build();
-    }
-
-
-    private SectionNode buildSection(Context context) {
-
-        return SectionNode.builder().tableOfContents(context.tableOfContents()).build();
-
-    }
-
-
-    private ParagraphNode buildParagraph(Context context, boolean terminal) {
-
-        return ParagraphNode.builder().terminal(terminal).tableOfContents(context.tableOfContents()).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 PageNode buildPage(PageData p) {
-
-        return PageNode.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 PageNode 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,
-            TableOfContents tableOfContents,
-            List<PageNode> pages,
-            List<SectionNode> sections,
-            List<AtomicTextBlockData> atomicTextBlockData,
-            List<AtomicPositionBlockData> atomicPositionBlockData) {
-
-    }
-
-}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/mapper/PropertiesMapper.java

@@ -1,101 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.mapper;
-
-import java.awt.geom.Rectangle2D;
-import java.util.Arrays;
-import java.util.HashMap;
-import java.util.List;
-import java.util.Map;
-
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.ImageNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.ImageType;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.TableCellNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.TableNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.services.RectangleTransformations;
-
-public class PropertiesMapper {
-
-    public static Map<String, String> buildImageProperties(ImageNode image) {
-
-        Map<String, String> properties = new HashMap<>();
-        properties.put("imageType", image.getImageType().toString());
-        properties.put("transparency", String.valueOf(image.isTransparency()));
-        properties.put("position", RectangleTransformations.toString(image.getPosition()));
-        return properties;
-    }
-
-
-    public static Map<String, String> buildTableCellProperties(TableCellNode 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 = RectangleTransformations.toString(tableCell.getBBox().get(tableCell.getPages().stream().findFirst().get()));
-        properties.put("bBox", bBoxString);
-
-        return properties;
-    }
-
-
-    public static Map<String, String> buildTableProperties(TableNode 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 void parseImageProperties(Map<String, String> properties, ImageNode.ImageNodeBuilder builder) {
-
-        builder.imageType(parseImageType(properties.get("imageType")));
-        builder.transparency(Boolean.parseBoolean(properties.get("transparency")));
-        builder.position(parseRectangle2D(properties.get("position")));
-    }
-
-
-    public static void parseTableCellProperties(Map<String, String> properties, TableCellNode.TableCellNodeBuilder 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, TableNode.TableNodeBuilder builder) {
-
-        builder.numberOfRows(Integer.parseInt(properties.get("numberOfRows")));
-        builder.numberOfCols(Integer.parseInt(properties.get("numberOfCols")));
-    }
-
-
-    private static ImageType parseImageType(String imageType) {
-
-        return switch (imageType) {
-            case "LOGO" -> ImageType.LOGO;
-            case "FORMULA" -> ImageType.FORMULA;
-            case "SIGNATURE" -> ImageType.SIGNATURE;
-            case "OCR" -> ImageType.OCR;
-            default -> ImageType.OTHER;
-        };
-    }
-
-
-    public static String toString(Rectangle2D rectangle2D) {
-
-        return String.format("%f,%f,%f,%f", rectangle2D.getX(), rectangle2D.getY(), rectangle2D.getWidth(), rectangle2D.getHeight());
-    }
-
-
-    public static Rectangle2D parseRectangle2D(String bBox) {
-
-        List<Float> floats = Arrays.stream(bBox.split(",")).map(Float::parseFloat).toList();
-        return new Rectangle2D.Float(floats.get(0), floats.get(1), floats.get(2), floats.get(3));
-    }
-
-}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/queue/LayoutParsingQueueNames.java

@@ -2,7 +2,7 @@ package com.knecon.fforesight.service.layoutparser.internal.api.queue;
 
 public class LayoutParsingQueueNames {
 
-    public static final String LAYOUT_PARSING_REQUEST_QUEUE = "LAYOUTPARSING_REQUEST_QUEUE";
-    public static final String LAYOUT_PARSING_DLQ = "LAYOUTPARSING_DLQ";
-    public static final String LAYOUT_PARSING_FINISHED_EVENT_QUEUE = "LAYOUTPARSING_FINISHED_EVENT_QUEUE";
+    public static final String LAYOUT_PARSING_REQUEST_QUEUE = "layout_parsing_request_queue";
+    public static final String LAYOUT_PARSING_DLQ = "layout_parsing_dead_letter_queue";
+    public static final String LAYOUT_PARSING_FINISHED_EVENT_QUEUE = "layout_parsing_response_queue";
 }

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/queue/LayoutParsingRequest.java

@@ -13,6 +13,7 @@ public record LayoutParsingRequest(
         Optional<String> tablesFileStorageId,
         Optional<String> imagesFileStorageId,
         String structureFileStorageId,
+        String researchDocumentStorageId,
         String textBlockFileStorageId,
         String positionBlockFileStorageId,
         String pageFileStorageId) {

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/services/EntityEnrichmentService.java

@@ -1,10 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.services;
-
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.entity.EntityNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.textblock.TextBlock;
-
-public interface EntityEnrichmentService {
-
-    void enrichEntity(EntityNode entity, TextBlock textBlock);
-
-}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/services/EntityInsertionService.java

@@ -1,56 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.services;
-
-import java.util.Collections;
-import java.util.NoSuchElementException;
-import java.util.Set;
-
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.TableOfContents;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.entity.EntityNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.PageNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.SemanticNode;
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.textblock.TextBlock;
-
-import lombok.RequiredArgsConstructor;
-
-@RequiredArgsConstructor
-public class EntityInsertionService {
-
-    private final EntityEnrichmentService entityEnrichmentService;
-
-
-    public void addEntityToGraph(EntityNode entity, TableOfContents tableOfContents) {
-
-        try {
-            SemanticNode containingNode = tableOfContents.streamChildrenNodes(Collections.emptyList())
-                    .filter(node -> node.buildTextBlock().containsBoundary(entity.getBoundary()))
-                    .findFirst()
-                    .orElseThrow(() -> new NoSuchElementException("No containing Node found!"));
-
-            containingNode.addThisToEntityIfIntersects(entity);
-
-            TextBlock textBlock = entity.getDeepestFullyContainingNode().buildTextBlock();
-            entityEnrichmentService.enrichEntity(entity, textBlock);
-
-            addToPages(entity);
-            addToNodeEntitySets(entity);
-
-        } catch (NoSuchElementException e) {
-            entity.removeFromGraph();
-        }
-    }
-
-
-    private void addToPages(EntityNode entity) {
-
-        Set<PageNode> pages = entity.getDeepestFullyContainingNode().getPages();
-        entity.getPages().addAll(pages);
-        pages.forEach(page -> page.getEntities().add(entity));
-    }
-
-
-    private void addToNodeEntitySets(EntityNode entity) {
-
-        entity.getIntersectingNodes().forEach(node -> node.getEntities().add(entity));
-    }
-
-}

layoutparser-service/layoutparser-service-processor/pom.xml

@@ -6,7 +6,7 @@
     <parent>
         <groupId>com.knecon.fforesight</groupId>
         <artifactId>layoutparser-service</artifactId>
-        <version>1.0.0</version>
+        <version>0.1-SNAPSHOT</version>
     </parent>
 
     <artifactId>layoutparser-service-processor</artifactId>
@@ -14,9 +14,14 @@
     <dependencies>
         <dependency>
             <groupId>com.iqser.red.service</groupId>
-            <artifactId>persistence-service-internal-api-v1</artifactId>
+            <artifactId>persistence-service-shared-api-v1</artifactId>
             <version>2.36.0</version>
         </dependency>
+        <dependency>
+            <groupId>com.knecon.fforesight</groupId>
+            <artifactId>tenant-commons</artifactId>
+            <version>${tennat-commons.version}</version>
+        </dependency>
         <dependency>
             <groupId>com.knecon.fforesight</groupId>
             <artifactId>layoutparser-service-internal-api</artifactId>
@@ -25,17 +30,7 @@
         <dependency>
             <groupId>com.iqser.red.commons</groupId>
             <artifactId>storage-commons</artifactId>
-            <version>1.13.0</version>
-        </dependency>
-        <dependency>
-            <groupId>com.iqser.red.commons</groupId>
-            <artifactId>spring-commons</artifactId>
-            <version>6.2.0</version>
-        </dependency>
-        <dependency>
-            <groupId>com.dslplatform</groupId>
-            <artifactId>dsl-json-java8</artifactId>
-            <version>1.10.0</version>
+            <version>${storage-commons.version}</version>
         </dependency>
         <dependency>
             <groupId>org.apache.pdfbox</groupId>
@@ -47,11 +42,6 @@
             <artifactId>pdfbox-tools</artifactId>
             <version>${pdfbox.version}</version>
         </dependency>
-        <dependency>
-            <groupId>com.google.guava</groupId>
-            <artifactId>guava</artifactId>
-            <version>31.1-jre</version>
-        </dependency>
         <dependency>
             <groupId>com.fasterxml.jackson.module</groupId>
             <artifactId>jackson-module-afterburner</artifactId>
@@ -62,68 +52,14 @@
             <artifactId>jackson-datatype-jsr310</artifactId>
             <version>${jackson.version}</version>
         </dependency>
-        <dependency>
-            <groupId>org.springframework.boot</groupId>
-            <artifactId>spring-boot-starter-security</artifactId>
-            <version>${spring.version}</version>
-        </dependency>
         <dependency>
             <groupId>org.springframework.boot</groupId>
             <artifactId>spring-boot-starter-web</artifactId>
-            <version>${spring.version}</version>
-        </dependency>
-        <dependency>
-            <groupId>org.projectlombok</groupId>
-            <artifactId>lombok</artifactId>
-            <optional>true</optional>
-        </dependency>
-        <dependency>
-            <groupId>org.springframework.cloud</groupId>
-            <artifactId>spring-cloud-starter-openfeign</artifactId>
-            <version>4.0.2</version>
         </dependency>
         <dependency>
             <groupId>org.springframework.boot</groupId>
             <artifactId>spring-boot-starter-amqp</artifactId>
-            <version>${spring.version}</version>
         </dependency>
     </dependencies>
 
-    <repositories>
-        <repository>
-            <id>spring-milestones</id>
-            <name>Spring Milestones</name>
-            <url>https://repo.spring.io/milestone</url>
-            <snapshots>
-                <enabled>false</enabled>
-            </snapshots>
-        </repository>
-        <repository>
-            <id>spring-snapshots</id>
-            <name>Spring Snapshots</name>
-            <url>https://repo.spring.io/snapshot</url>
-            <releases>
-                <enabled>false</enabled>
-            </releases>
-        </repository>
-    </repositories>
-    <pluginRepositories>
-        <pluginRepository>
-            <id>spring-milestones</id>
-            <name>Spring Milestones</name>
-            <url>https://repo.spring.io/milestone</url>
-            <snapshots>
-                <enabled>false</enabled>
-            </snapshots>
-        </pluginRepository>
-        <pluginRepository>
-            <id>spring-snapshots</id>
-            <name>Spring Snapshots</name>
-            <url>https://repo.spring.io/snapshot</url>
-            <releases>
-                <enabled>false</enabled>
-            </releases>
-        </pluginRepository>
-    </pluginRepositories>
-
 </project>

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/LayoutParsingService.java

@@ -7,19 +7,21 @@ import java.io.IOException;
 import org.apache.pdfbox.pdmodel.PDDocument;
 import org.springframework.stereotype.Service;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.DocumentGraph;
-import com.knecon.fforesight.service.layoutparser.internal.api.mapper.DocumentDataMapper;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.taas.ResearchDocumentData;
+import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingFinishedEvent;
+import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingRequest;
 import com.knecon.fforesight.service.layoutparser.processor.adapter.CvTableParsingAdapter;
 import com.knecon.fforesight.service.layoutparser.processor.adapter.ImageServiceResponseAdapter;
 import com.knecon.fforesight.service.layoutparser.processor.adapter.model.image.ImageServiceResponse;
 import com.knecon.fforesight.service.layoutparser.processor.adapter.model.table.TableServiceResponse;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.ClassificationDocument;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.ClassificationDocument;
 import com.knecon.fforesight.service.layoutparser.processor.classification.service.ClassificationService;
 import com.knecon.fforesight.service.layoutparser.processor.classification.service.PdfParsingService;
 import com.knecon.fforesight.service.layoutparser.processor.classification.service.SectionsBuilderService;
 import com.knecon.fforesight.service.layoutparser.processor.factory.DocumentGraphFactory;
-import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingFinishedEvent;
-import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingRequest;
+import com.knecon.fforesight.service.layoutparser.processor.graph.nodes.Document;
+import com.knecon.fforesight.service.layoutparser.processor.mapper.redaction.DocumentDataMapper;
+import com.knecon.fforesight.service.layoutparser.processor.mapper.taas.TaasDocumentDataMapper;
 
 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;
@@ -35,7 +37,6 @@ public class LayoutParsingService {
     private final PdfParsingService pdfParsingService;
     private final ClassificationService classificationService;
     private final SectionsBuilderService sectionsBuilderService;
-    private final DocumentGraphFactory documentGraphFactory;
 
 
     public LayoutParsingFinishedEvent parseLayoutAndSaveFilesToStorage(LayoutParsingRequest layoutParsingRequest) throws IOException {
@@ -53,11 +54,13 @@ public class LayoutParsingService {
             tableServiceResponse = layoutParsingStorageService.getTablesFile(layoutParsingRequest.pageFileStorageId());
         }
 
-        DocumentGraph documentGraph = parseLayout(originDocument, imageServiceResponse, tableServiceResponse);
+        Document documentGraph = parseLayout(originDocument, imageServiceResponse, tableServiceResponse);
         int numberOfPages = originDocument.getNumberOfPages();
         originDocument.close();
 
-        layoutParsingStorageService.storeDocumentData(layoutParsingRequest, DocumentDataMapper.toDocumentData(documentGraph));
+        var researchDocumentData = TaasDocumentDataMapper.fromDocument(documentGraph);
+
+        layoutParsingStorageService.storeDocumentData(layoutParsingRequest, researchDocumentData, DocumentDataMapper.toDocumentData(documentGraph));
 
         return LayoutParsingFinishedEvent.builder()
                 .identifier(layoutParsingRequest.identifier())
@@ -72,7 +75,7 @@ public class LayoutParsingService {
     }
 
 
-    public DocumentGraph parseLayout(PDDocument originDocument, ImageServiceResponse imageServiceResponse, TableServiceResponse tableServiceResponse) {
+    public Document parseLayout(PDDocument originDocument, ImageServiceResponse imageServiceResponse, TableServiceResponse tableServiceResponse) {
 
         ClassificationDocument classificationDocument = pdfParsingService.parseDocument(originDocument,
                 cvTableParsingAdapter.buildCvParsedTablesPerPage(tableServiceResponse),
@@ -82,7 +85,30 @@ public class LayoutParsingService {
 
         sectionsBuilderService.buildSections(classificationDocument);
 
-        return documentGraphFactory.buildDocumentGraph(classificationDocument);
+        return DocumentGraphFactory.buildDocumentGraph(classificationDocument);
+    }
+
+
+    public Document parseLayoutWithTimer(PDDocument originDocument, ImageServiceResponse imageServiceResponse, TableServiceResponse tableServiceResponse) {
+
+        long start = System.currentTimeMillis();
+        ClassificationDocument classificationDocument = pdfParsingService.parseDocument(originDocument,
+                cvTableParsingAdapter.buildCvParsedTablesPerPage(tableServiceResponse),
+                imageServiceResponseAdapter.buildClassifiedImagesPerPage(imageServiceResponse));
+        System.out.printf("parsed %d ms", System.currentTimeMillis() - start);
+
+        start = System.currentTimeMillis();
+        classificationService.classifyDocument(classificationDocument);
+        System.out.printf(", classified %d ms", System.currentTimeMillis() - start);
+
+        start = System.currentTimeMillis();
+        sectionsBuilderService.buildSections(classificationDocument);
+        System.out.printf(", sections built %d ms", System.currentTimeMillis() - start);
+
+        start = System.currentTimeMillis();
+        Document document = DocumentGraphFactory.buildDocumentGraph(classificationDocument);
+        System.out.printf(", graph constructed %d ms", System.currentTimeMillis() - start);
+        return document;
     }
 
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/LayoutParsingStorageService.java

@@ -14,15 +14,16 @@ import org.springframework.stereotype.Service;
 
 import com.fasterxml.jackson.databind.ObjectMapper;
 import com.iqser.red.storage.commons.service.StorageService;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.AtomicPositionBlockData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.AtomicTextBlockData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.DocumentData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.PageData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.TableOfContentsData;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.AtomicPositionBlockData;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.AtomicTextBlockData;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentData;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentTreeData;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.PageData;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.taas.ResearchDocumentData;
+import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingRequest;
 import com.knecon.fforesight.service.layoutparser.processor.adapter.model.image.ImageServiceResponse;
 import com.knecon.fforesight.service.layoutparser.processor.adapter.model.table.TableServiceResponse;
-import com.knecon.fforesight.service.layoutparser.processor.multitenancy.TenantContext;
-import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingRequest;
+import com.knecon.fforesight.tenantcommons.TenantContext;
 
 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;
@@ -67,9 +68,10 @@ public class LayoutParsingStorageService {
     }
 
 
-    public void storeDocumentData(LayoutParsingRequest layoutParsingRequest, DocumentData documentData) throws IOException {
+    public void storeDocumentData(LayoutParsingRequest layoutParsingRequest, ResearchDocumentData researchDocumentData, DocumentData documentData)  {
 
-        storageService.storeJSONObject(TenantContext.getTenantId(), layoutParsingRequest.structureFileStorageId(), documentData.getTableOfContents());
+        storageService.storeJSONObject(TenantContext.getTenantId(), layoutParsingRequest.researchDocumentStorageId(), researchDocumentData);
+        storageService.storeJSONObject(TenantContext.getTenantId(), layoutParsingRequest.structureFileStorageId(), documentData.getDocumentTreeData());
         storageService.storeJSONObject(TenantContext.getTenantId(), layoutParsingRequest.textBlockFileStorageId(), documentData.getAtomicTextBlocks());
         storageService.storeJSONObject(TenantContext.getTenantId(), layoutParsingRequest.positionBlockFileStorageId(), documentData.getAtomicPositionBlocks());
         storageService.storeJSONObject(TenantContext.getTenantId(), layoutParsingRequest.pageFileStorageId(), documentData.getPages());
@@ -86,12 +88,12 @@ public class LayoutParsingStorageService {
         AtomicPositionBlockData[] atomicPositionBlockData = storageService.readJSONObject(TenantContext.getTenantId(),
                 layoutParsingRequest.positionBlockFileStorageId(),
                 AtomicPositionBlockData[].class);
-        TableOfContentsData tableOfContentsData = storageService.readJSONObject(TenantContext.getTenantId(),
+        DocumentTreeData tableOfContentsData = storageService.readJSONObject(TenantContext.getTenantId(),
                 layoutParsingRequest.structureFileStorageId(),
-                TableOfContentsData.class);
+                DocumentTreeData.class);
 
         return DocumentData.builder()
-                .tableOfContents(tableOfContentsData)
+                .documentTreeData(tableOfContentsData)
                 .atomicPositionBlocks(atomicPositionBlockData)
                 .atomicTextBlocks(atomicTextBlockData)
                 .pages(pageData)

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/adapter/CvTableParsingAdapter.java

@@ -8,7 +8,7 @@ import java.util.Map;
 
 import org.springframework.stereotype.Service;
 
-import com.knecon.fforesight.service.layoutparser.processor.adapter.model.table.CvParsedTableCell;
+import com.knecon.fforesight.service.layoutparser.processor.adapter.model.table.TableCells;
 import com.knecon.fforesight.service.layoutparser.processor.adapter.model.table.TableServiceResponse;
 
 import lombok.RequiredArgsConstructor;
@@ -19,9 +19,9 @@ import lombok.extern.slf4j.Slf4j;
 @RequiredArgsConstructor
 public class CvTableParsingAdapter {
 
-    public Map<Integer, List<com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.CvParsedTableCell>> buildCvParsedTablesPerPage(TableServiceResponse tableServiceResponse) {
+    public Map<Integer, List<TableCells>> buildCvParsedTablesPerPage(TableServiceResponse tableServiceResponse) {
 
-        Map<Integer, List<com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.CvParsedTableCell>> tableCells = new HashMap<>();
+        Map<Integer, List<TableCells>> tableCells = new HashMap<>();
         tableServiceResponse.getData()
                 .forEach(tableData -> tableCells.computeIfAbsent(tableData.getPageInfo().getNumber(), tableCell -> new ArrayList<>())
                         .addAll(convertTableCells(tableData.getTableCells())));
@@ -30,11 +30,11 @@ public class CvTableParsingAdapter {
     }
 
 
-    private Collection<? extends com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.CvParsedTableCell> convertTableCells(List<CvParsedTableCell> tableCells) {
+    private Collection<TableCells> convertTableCells(List<TableCells> tableCells) {
 
-        List<com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.CvParsedTableCell> cvParsedTableCells = new ArrayList<>();
+        List<TableCells> cvParsedTableCells = new ArrayList<>();
 
-        tableCells.forEach(t -> cvParsedTableCells.add(com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.CvParsedTableCell.builder()
+        tableCells.forEach(t -> cvParsedTableCells.add(TableCells.builder()
                 .y0(t.getY0())
                 .x1(t.getX1())
                 .y1(t.getY1())

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/adapter/ImageServiceResponseAdapter.java

@@ -9,10 +9,10 @@ import java.util.Map;
 
 import org.springframework.stereotype.Service;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.ImageType;
 import com.knecon.fforesight.service.layoutparser.processor.adapter.model.image.ImageServiceResponse;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.ClassificationPage;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.image.ClassifiedImage;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.ClassificationPage;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.image.ClassifiedImage;
+import com.knecon.fforesight.service.layoutparser.processor.graph.nodes.ImageType;
 
 import lombok.RequiredArgsConstructor;
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/adapter/model/image/Classification.java

@@ -3,12 +3,9 @@ package com.knecon.fforesight.service.layoutparser.processor.adapter.model.image
 import java.util.HashMap;
 import java.util.Map;
 
-import com.dslplatform.json.CompiledJson;
-
 import lombok.Data;
 
 @Data
-@CompiledJson
 public class Classification {
 
     private Map<String, Float> probabilities = new HashMap<>();

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/adapter/model/image/FilterGeometry.java

@@ -1,14 +1,11 @@
 package com.knecon.fforesight.service.layoutparser.processor.adapter.model.image;
 
-import com.dslplatform.json.CompiledJson;
-
 import lombok.Data;
 
 @Data
-@CompiledJson
 public class FilterGeometry {
 
     private ImageSize imageSize;
-    private Format imageFormat;
+    private ImageFormat imageFormat;
 
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/adapter/model/image/Filters.java

@@ -1,11 +1,8 @@
 package com.knecon.fforesight.service.layoutparser.processor.adapter.model.image;
 
-import com.dslplatform.json.CompiledJson;
-
 import lombok.Data;
 
 @Data
-@CompiledJson
 public class Filters {
 
     private FilterGeometry geometry;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/adapter/model/image/Geometry.java

@@ -1,11 +1,8 @@
 package com.knecon.fforesight.service.layoutparser.processor.adapter.model.image;
 
-import com.dslplatform.json.CompiledJson;
-
 import lombok.Data;
 
 @Data
-@CompiledJson
 public class Geometry {
 
     private float width;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/adapter/model/image/ImageFormat.java

@@ -1,12 +1,9 @@
 package com.knecon.fforesight.service.layoutparser.processor.adapter.model.image;
 
-import com.dslplatform.json.CompiledJson;
-
 import lombok.Data;
 
 @Data
-@CompiledJson
-public class Format {
+public class ImageFormat {
 
     private float quotient;
     private boolean tooTall;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/adapter/model/image/ImageMetadata.java

@@ -1,12 +1,9 @@
 package com.knecon.fforesight.service.layoutparser.processor.adapter.model.image;
 
-import com.dslplatform.json.CompiledJson;
-
 import lombok.Data;
 
 @Data
-@CompiledJson
-public class Metadata {
+public class ImageMetadata {
 
     private Classification classification;
     private Position position;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/adapter/model/image/ImageServiceResponse.java

@@ -3,15 +3,12 @@ package com.knecon.fforesight.service.layoutparser.processor.adapter.model.image
 import java.util.ArrayList;
 import java.util.List;
 
-import com.dslplatform.json.CompiledJson;
-import com.dslplatform.json.JsonAttribute;
 import com.fasterxml.jackson.annotation.JsonAlias;
 import com.fasterxml.jackson.annotation.JsonProperty;
 
 import lombok.Data;
 
 @Data
-@CompiledJson
 public class ImageServiceResponse {
 
     private String dossierId;
@@ -19,15 +16,13 @@ public class ImageServiceResponse {
 
     @JsonProperty(value = "imageMetadata")
     @JsonAlias("data")
-    @JsonAttribute(alternativeNames = {"imageMetadata"})
-    private List<Metadata> data = new ArrayList<>();
+    private List<ImageMetadata> data = new ArrayList<>();
 
-    private List<Metadata> dataCV = new ArrayList<>();
+    private List<ImageMetadata> dataCV = new ArrayList<>();
 
 
     @JsonProperty(value = "imageMetadata")
     @JsonAlias("data")
-    @JsonAttribute(alternativeNames = {"imageMetadata"})
-    public void setData(List<Metadata> data) {this.data = data;}
+    public void setData(List<ImageMetadata> data) {this.data = data;}
 
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/adapter/model/image/ImageSize.java

@@ -1,11 +1,8 @@
 package com.knecon.fforesight.service.layoutparser.processor.adapter.model.image;
 
-import com.dslplatform.json.CompiledJson;
-
 import lombok.Data;
 
 @Data
-@CompiledJson
 public class ImageSize {
 
     private float quotient;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/adapter/model/image/Position.java

@@ -1,11 +1,8 @@
 package com.knecon.fforesight.service.layoutparser.processor.adapter.model.image;
 
-import com.dslplatform.json.CompiledJson;
-
 import lombok.Data;
 
 @Data
-@CompiledJson
 public class Position {
 
     private float x1;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/adapter/model/image/Probability.java

@@ -1,11 +1,8 @@
 package com.knecon.fforesight.service.layoutparser.processor.adapter.model.image;
 
-import com.dslplatform.json.CompiledJson;
-
 import lombok.Data;
 
 @Data
-@CompiledJson
 public class Probability {
 
     private boolean unconfident;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/adapter/model/table/CvParsedTableModel.java

@@ -1,17 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.processor.adapter.model.table;
-
-import java.util.ArrayList;
-import java.util.List;
-
-import com.dslplatform.json.CompiledJson;
-
-import lombok.Data;
-
-@Data
-@CompiledJson
-public class CvParsedTableModel {
-
-    private CvParsedPageInfo pageInfo;
-    private List<CvParsedTableCell> tableCells = new ArrayList<>();
-
-}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/adapter/model/table/PageInfo.java

@@ -1,12 +1,9 @@
 package com.knecon.fforesight.service.layoutparser.processor.adapter.model.table;
 
-import com.dslplatform.json.CompiledJson;
-
 import lombok.Data;
 
 @Data
-@CompiledJson
-public class CvParsedPageInfo {
+public class PageInfo {
 
     private int number;
     private int rotation;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/adapter/model/table/PdfTableCell.java

@@ -1,4 +1,4 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto.table;
+package com.knecon.fforesight.service.layoutparser.processor.adapter.model.table;
 
 import lombok.AllArgsConstructor;
 import lombok.Builder;
@@ -9,7 +9,7 @@ import lombok.RequiredArgsConstructor;
 @Builder
 @AllArgsConstructor
 @RequiredArgsConstructor
-public class CvParsedTableCell {
+public class PdfTableCell {
 
     private float x0;
     private float y0;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/adapter/model/table/TableCells.java

@@ -1,12 +1,11 @@
 package com.knecon.fforesight.service.layoutparser.processor.adapter.model.table;
 
-import com.dslplatform.json.CompiledJson;
-
+import lombok.Builder;
 import lombok.Data;
 
 @Data
-@CompiledJson
-public class CvParsedTableCell {
+@Builder
+public class TableCells {
 
     private float x0;
     private float y0;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/adapter/model/table/TableData.java

@@ -0,0 +1,14 @@
+package com.knecon.fforesight.service.layoutparser.processor.adapter.model.table;
+
+import java.util.ArrayList;
+import java.util.List;
+
+import lombok.Data;
+
+@Data
+public class TableData {
+
+    private PageInfo pageInfo;
+    private List<TableCells> tableCells = new ArrayList<>();
+
+}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/adapter/model/table/TableServiceResponse.java

@@ -3,12 +3,9 @@ package com.knecon.fforesight.service.layoutparser.processor.adapter.model.table
 import java.util.ArrayList;
 import java.util.List;
 
-import com.dslplatform.json.CompiledJson;
-
 import lombok.Data;
 
 @Data
-@CompiledJson
 public class TableServiceResponse {
 
     private String dossierId;
@@ -17,6 +14,6 @@ public class TableServiceResponse {
     private String targetFileExtension;
     private String responseFileExtension;
 
-    private List<CvParsedTableModel> data = new ArrayList<>();
+    private List<TableData> data = new ArrayList<>();
 
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/dto/AbstractTextContainer.java

@@ -1,71 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto;
-
-import java.awt.geom.Rectangle2D;
-
-import com.dslplatform.json.JsonAttribute;
-import com.fasterxml.jackson.annotation.JsonIgnore;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.ClassificationTextBlock;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.TextBlockOrientation;
-
-import lombok.AllArgsConstructor;
-import lombok.Data;
-import lombok.NoArgsConstructor;
-
-@Data
-@AllArgsConstructor
-@NoArgsConstructor
-public abstract class AbstractTextContainer {
-
-    protected float minX;
-    protected float maxX;
-    protected float minY;
-    protected float maxY;
-    protected String classification;
-    protected int page;
-
-    private TextBlockOrientation orientation = TextBlockOrientation.NONE;
-
-
-    public abstract String getText();
-
-
-    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) {
-
-        return this.minX <= other.minX && this.maxX >= other.maxX && this.minY >= other.minY && this.maxY <= other.maxY;
-    }
-
-
-    public boolean contains(Rectangle2D other) {
-
-        return other.contains(minX, minY, getWidth(), getHeight());
-    }
-
-
-    @JsonIgnore
-    @JsonAttribute(ignore = true)
-    public float getHeight() {
-
-        return maxY - minY;
-    }
-
-
-    @JsonIgnore
-    @JsonAttribute(ignore = true)
-    public float getWidth() {
-
-        return maxX - minX;
-    }
-
-
-    public boolean intersectsY(AbstractTextContainer atc) {
-
-        return this.minY <= atc.getMaxY() && this.maxY >= atc.getMinY();
-    }
-
-}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/dto/table/TableCell.java

@@ -1,38 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto.table;
-
-import java.awt.geom.Point2D;
-import java.util.ArrayList;
-import java.util.List;
-
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.ClassificationTextBlock;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.Rectangle;
-
-import lombok.Data;
-import lombok.EqualsAndHashCode;
-import lombok.NoArgsConstructor;
-
-@SuppressWarnings("serial")
-@Data
-@EqualsAndHashCode(callSuper = true)
-@NoArgsConstructor
-public class TableCell extends Rectangle {
-
-    private List<ClassificationTextBlock> textBlocks = new ArrayList<>();
-
-    private List<TableCell> headerCells = new ArrayList<>();
-
-    private boolean isHeaderCell;
-
-
-    public TableCell(Point2D topLeft, Point2D bottomRight) {
-
-        super((float) topLeft.getY(), (float) topLeft.getX(), (float) (bottomRight.getX() - topLeft.getX()), (float) (bottomRight.getY() - topLeft.getY()));
-    }
-
-
-    public void addTextBlock(ClassificationTextBlock textBlock) {
-
-        textBlocks.add(textBlock);
-    }
-
-}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/AbstractPageBlock.java

@@ -0,0 +1,80 @@
+package com.knecon.fforesight.service.layoutparser.processor.classification.model;
+
+import com.fasterxml.jackson.annotation.JsonIgnore;
+import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.Rectangle;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.text.TextPageBlock;
+
+import lombok.AllArgsConstructor;
+import lombok.Data;
+import lombok.NoArgsConstructor;
+
+@Data
+@AllArgsConstructor
+@NoArgsConstructor
+public abstract class AbstractPageBlock {
+
+    @JsonIgnore
+    protected float minX;
+    @JsonIgnore
+    protected float maxX;
+    @JsonIgnore
+    protected float minY;
+    @JsonIgnore
+    protected float maxY;
+    @JsonIgnore
+    protected PageBlockType classification;
+    @JsonIgnore
+    protected int page;
+
+    @JsonIgnore
+    private Orientation orientation = Orientation.NONE;
+
+
+    public abstract String getText();
+
+
+    public boolean isHeadline() {
+
+        return this instanceof TextPageBlock && this.getClassification() != null && this.getClassification().isHeadline();
+    }
+
+
+    public boolean containsBlock(TextPageBlock other) {
+
+        return this.minX <= other.getPdfMinX() && this.maxX >= other.getPdfMaxX() && this.minY >= other.getPdfMinY() && this.maxY <= other.getPdfMaxY();
+    }
+
+
+    public boolean contains(AbstractPageBlock other) {
+
+        return this.minX <= other.minX && this.maxX >= other.maxX && this.minY >= other.minY && this.maxY <= other.maxY;
+    }
+
+
+    public boolean contains(Rectangle other) {
+
+        return page == other.getPage() && this.minX <= other.getTopLeft().getX() && this.maxX >= other.getTopLeft().getX() + other.getWidth() && this.minY <= other.getTopLeft()
+                .getY() && this.maxY >= other.getTopLeft().getY() + other.getHeight();
+    }
+
+
+    @JsonIgnore
+    public float getHeight() {
+
+        return maxY - minY;
+    }
+
+
+    @JsonIgnore
+    public float getWidth() {
+
+        return maxX - minX;
+    }
+
+
+    public boolean intersectsY(AbstractPageBlock atc) {
+
+        return this.minY <= atc.getMaxY() && this.maxY >= atc.getMinY();
+    }
+
+}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/ClassificationDocument.java

@@ -1,10 +1,11 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto;
+package com.knecon.fforesight.service.layoutparser.processor.classification.model;
 
 import java.util.ArrayList;
 import java.util.List;
 
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.StringFrequencyCounter;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.UnclassifiedText;
+import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.section.SectionGrid;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.text.StringFrequencyCounter;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.text.UnclassifiedText;
 
 import lombok.Data;
 import lombok.NoArgsConstructor;
@@ -24,4 +25,7 @@ public class ClassificationDocument {
     private StringFrequencyCounter fontStyleCounter = new StringFrequencyCounter();
     private boolean headlines;
 
+    private SectionGrid sectionGrid = new SectionGrid();
+    private long rulesVersion;
+
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/ClassificationFooter.java

@@ -1,8 +1,8 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto;
+package com.knecon.fforesight.service.layoutparser.processor.classification.model;
 
 import java.util.List;
 
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.ClassificationTextBlock;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.text.TextPageBlock;
 
 import lombok.AllArgsConstructor;
 import lombok.Data;
@@ -11,6 +11,6 @@ import lombok.Data;
 @AllArgsConstructor
 public class ClassificationFooter {
 
-    private List<ClassificationTextBlock> textBlocks;
+    private List<TextPageBlock> textBlocks;
 
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/ClassificationHeader.java

@@ -1,8 +1,8 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto;
+package com.knecon.fforesight.service.layoutparser.processor.classification.model;
 
 import java.util.List;
 
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.ClassificationTextBlock;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.text.TextPageBlock;
 
 import lombok.AllArgsConstructor;
 import lombok.Data;
@@ -11,6 +11,6 @@ import lombok.Data;
 @AllArgsConstructor
 public class ClassificationHeader {
 
-    private List<ClassificationTextBlock> textBlocks;
+    private List<TextPageBlock> textBlocks;
 
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/ClassificationPage.java

@@ -1,11 +1,11 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto;
+package com.knecon.fforesight.service.layoutparser.processor.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.knecon.fforesight.service.layoutparser.processor.classification.dto.image.ClassifiedImage;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.StringFrequencyCounter;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.image.ClassifiedImage;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.text.StringFrequencyCounter;
 
 import lombok.Data;
 import lombok.NonNull;
@@ -16,7 +16,7 @@ import lombok.RequiredArgsConstructor;
 public class ClassificationPage {
 
     @NonNull
-    private List<AbstractTextContainer> textBlocks;
+    private List<AbstractPageBlock> textBlocks;
 
     private List<ClassifiedImage> images = new ArrayList<>();
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/ClassificationSection.java

@@ -1,38 +1,32 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto;
+package com.knecon.fforesight.service.layoutparser.processor.classification.model;
 
 import java.util.ArrayList;
 import java.util.List;
 
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.image.ClassifiedImage;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.Table;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.image.ClassifiedImage;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.table.TablePageBlock;
 
 import lombok.Data;
 import lombok.NoArgsConstructor;
 
 @Data
 @NoArgsConstructor
-public class ClassificationSection implements Comparable {
+public class ClassificationSection {
 
-    private List<AbstractTextContainer> pageBlocks = new ArrayList<>();
+    private List<AbstractPageBlock> pageBlocks = new ArrayList<>();
     private List<ClassifiedImage> images = new ArrayList<>();
     private String headline;
 
 
-    public List<Table> getTables() {
+    public List<TablePageBlock> getTables() {
 
-        List<Table> tables = new ArrayList<>();
+        List<TablePageBlock> tables = new ArrayList<>();
         pageBlocks.forEach(block -> {
-            if (block instanceof Table) {
-                tables.add((Table) block);
+            if (block instanceof TablePageBlock) {
+                tables.add((TablePageBlock) block);
             }
         });
         return tables;
     }
 
-
-    @Override
-    public int compareTo(Object o) {
-
-        return 0;
-    }
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/FloatFrequencyCounter.java

@@ -1,4 +1,4 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto;
+package com.knecon.fforesight.service.layoutparser.processor.classification.model;
 
 import java.util.ArrayList;
 import java.util.Collections;
@@ -9,9 +9,9 @@ import java.util.stream.Collectors;
 
 import lombok.Getter;
 
+@Getter
 public class FloatFrequencyCounter {
 
-    @Getter
     Map<Float, Integer> countPerValue = new HashMap<>();
 
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/Orientation.java

@@ -1,6 +1,6 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto.text;
+package com.knecon.fforesight.service.layoutparser.processor.classification.model;
 
-public enum TextBlockOrientation {
+public enum Orientation {
 
     NONE,
     LEFT,

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/PageBlockType.java

@@ -0,0 +1,38 @@
+package com.knecon.fforesight.service.layoutparser.processor.classification.model;
+
+public enum PageBlockType {
+    H1,
+    H2,
+    H3,
+    H4,
+    H5,
+    H6,
+    HEADER,
+    FOOTER,
+    TITLE,
+    PARAGRAPH,
+    PARAGRAPH_BOLD,
+    PARAGRAPH_ITALIC,
+    PARAGRAPH_UNKNOWN,
+    OTHER,
+    TABLE;
+
+
+    public static PageBlockType getHeadlineType(int i) {
+
+        return switch (i) {
+            case 1 -> PageBlockType.H1;
+            case 2 -> PageBlockType.H2;
+            case 3 -> PageBlockType.H3;
+            case 4 -> PageBlockType.H4;
+            case 5 -> PageBlockType.H5;
+            default -> PageBlockType.H6;
+        };
+    }
+
+
+    public boolean isHeadline() {
+
+        return this.equals(H1) || this.equals(H2) || this.equals(H3) || this.equals(H4) || this.equals(H5) || this.equals(H6);
+    }
+}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/image/ClassifiedImage.java

@@ -1,8 +1,8 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto.image;
+package com.knecon.fforesight.service.layoutparser.processor.classification.model.image;
 
 import java.awt.geom.Rectangle2D;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.graph.nodes.ImageType;
+import com.knecon.fforesight.service.layoutparser.processor.graph.nodes.ImageType;
 
 import lombok.Data;
 import lombok.NonNull;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/table/Cell.java

@@ -0,0 +1,79 @@
+package com.knecon.fforesight.service.layoutparser.processor.classification.model.table;
+
+import java.awt.geom.Point2D;
+import java.util.ArrayList;
+import java.util.Iterator;
+import java.util.List;
+
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.text.TextPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.text.TextPositionSequence;
+import com.knecon.fforesight.service.layoutparser.processor.classification.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<TextPageBlock> textBlocks = new ArrayList<>();
+
+    private List<Cell> headerCells = new ArrayList<>();
+
+    private boolean isHeaderCell;
+
+    private static final int MIN_SIZE = 1;
+
+    private int pageNumber;
+
+
+    public Cell(Point2D topLeft, Point2D bottomRight) {
+
+        super((float) topLeft.getY(), (float) topLeft.getX(), (float) (bottomRight.getX() - topLeft.getX()), (float) (bottomRight.getY() - topLeft.getY()));
+    }
+
+
+    public void addTextBlock(TextPageBlock textBlock) {
+
+        textBlocks.add(textBlock);
+    }
+
+
+    @Override
+    public String toString() {
+
+        StringBuilder sb = new StringBuilder();
+
+        Iterator<TextPageBlock> itty = textBlocks.iterator();
+        TextPositionSequence previous = null;
+        while (itty.hasNext()) {
+
+            TextPageBlock textBlock = itty.next();
+
+            for (TextPositionSequence word : textBlock.getSequences()) {
+                if (previous != null) {
+                    if (Math.abs(previous.getMaxYDirAdj() - word.getMaxYDirAdj()) > word.getTextHeight()) {
+                        sb.append('\n');
+                    } else {
+                        sb.append(' ');
+                    }
+                }
+                sb.append(word.toString());
+                previous = word;
+            }
+
+        }
+
+        return TextNormalizationUtilities.removeHyphenLineBreaks(sb.toString()).replaceAll("\n", " ").replaceAll(" {2}", " ");
+    }
+
+
+    public boolean hasMinimumSize() {
+
+        return this.getHeight() >= MIN_SIZE && this.getWidth() >= MIN_SIZE;
+    }
+
+}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/table/CellPosition.java

@@ -1,11 +1,11 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto.table;
+package com.knecon.fforesight.service.layoutparser.processor.classification.model.table;
 
 import lombok.RequiredArgsConstructor;
 import lombok.Value;
 
 @Value
 @RequiredArgsConstructor
-public class TableCellPosition implements Comparable<TableCellPosition> {
+public class CellPosition implements Comparable<CellPosition> {
 
     int row;
 
@@ -13,7 +13,7 @@ public class TableCellPosition implements Comparable<TableCellPosition> {
 
 
     @Override
-    public int compareTo(TableCellPosition other) {
+    public int compareTo(CellPosition other) {
 
         int rowDiff = row - other.row;
         return rowDiff != 0 ? rowDiff : col - other.col;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/table/CleanRulings.java

@@ -1,4 +1,4 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto.table;
+package com.knecon.fforesight.service.layoutparser.processor.classification.model.table;
 
 import java.util.List;
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/table/Rectangle.java

@@ -1,4 +1,4 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto;
+package com.knecon.fforesight.service.layoutparser.processor.classification.model.table;
 
 import java.awt.geom.Point2D;
 import java.awt.geom.Rectangle2D;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/table/Ruling.java

@@ -1,4 +1,4 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto.table;
+package com.knecon.fforesight.service.layoutparser.processor.classification.model.table;
 
 import java.awt.geom.Line2D;
 import java.awt.geom.Point2D;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/table/TablePageBlock.java

@@ -1,4 +1,4 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto.table;
+package com.knecon.fforesight.service.layoutparser.processor.classification.model.table;
 
 import java.awt.geom.Point2D;
 import java.util.ArrayList;
@@ -7,20 +7,19 @@ import java.util.HashSet;
 import java.util.List;
 import java.util.Set;
 import java.util.TreeMap;
-import java.util.stream.Collectors;
 
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.AbstractTextContainer;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.ClassificationTextBlock;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.Rectangle;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.AbstractPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.PageBlockType;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.text.TextPageBlock;
 
 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<TableCellPosition, TableCell> cells = new TreeMap<>();
+    private final TreeMap<CellPosition, Cell> cells = new TreeMap<>();
 
     private final int rotation;
     @Getter
@@ -28,32 +27,29 @@ 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<TableCell>> rows;
+    private List<List<Cell>> rows;
 
 
-    public Table(List<TableCell> 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 = PageBlockType.TABLE;
         this.rotation = rotation;
-
     }
 
 
-    public List<List<TableCell>> getRows() {
+    public List<List<Cell>> getRows() {
 
         if (rows == null) {
             rows = computeRows();
 
             // Ignore rows that does not contain any cells and values.
-            List<List<TableCell>> rowsToRemove = new ArrayList<>();
-            for (List<TableCell> row : rows) {
+            List<List<Cell>> rowsToRemove = new ArrayList<>();
+            for (List<Cell> row : rows) {
                 if (row.size() == 1 && row.get(0).getTextBlocks().isEmpty()) {
                     rowsToRemove.add(row);
                 }
@@ -70,19 +66,13 @@ 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);
 
     }
 
@@ -100,16 +90,16 @@ public class Table extends AbstractTextContainer {
         // A bold cell is a header cell as long as every cell to the left/top is bold, too
         // we move from left to right and top to bottom
         for (int rowIndex = 0; rowIndex < rows.size(); rowIndex++) {
-            List<TableCell> rowCells = rows.get(rowIndex);
+            List<Cell> rowCells = rows.get(rowIndex);
             if (rowCells.size() == 1) {
                 continue;
             }
 
             for (int colIndex = 0; colIndex < rowCells.size(); colIndex++) {
-                TableCell cell = rowCells.get(colIndex);
-                List<TableCell> cellsToTheLeft = rowCells.subList(0, colIndex);
-                TableCell lastHeaderCell = null;
-                for (TableCell leftCell : cellsToTheLeft) {
+                Cell cell = rowCells.get(colIndex);
+                List<Cell> cellsToTheLeft = rowCells.subList(0, colIndex);
+                Cell lastHeaderCell = null;
+                for (Cell leftCell : cellsToTheLeft) {
                     if (leftCell.isHeaderCell()) {
                         lastHeaderCell = leftCell;
                     } else {
@@ -119,7 +109,7 @@ public class Table extends AbstractTextContainer {
                 if (lastHeaderCell != null) {
                     cell.getHeaderCells().add(lastHeaderCell);
                 }
-                List<TableCell> cellsToTheTop = new ArrayList<>();
+                List<Cell> cellsToTheTop = new ArrayList<>();
                 for (int i = 0; i < rowIndex; i++) {
                     try {
                         cellsToTheTop.add(rows.get(i).get(colIndex));
@@ -127,7 +117,7 @@ public class Table extends AbstractTextContainer {
                         log.debug("No cell {} in row {}, ignoring.", colIndex, rowIndex);
                     }
                 }
-                for (TableCell topCell : cellsToTheTop) {
+                for (Cell topCell : cellsToTheTop) {
                     if (topCell.isHeaderCell()) {
                         lastHeaderCell = topCell;
                     } else {
@@ -146,14 +136,14 @@ public class Table extends AbstractTextContainer {
     }
 
 
-    private List<List<TableCell>> computeRows() {
+    private List<List<Cell>> computeRows() {
 
-        List<List<TableCell>> rows = new ArrayList<>();
+        List<List<Cell>> rows = new ArrayList<>();
         if (rotation == 90) {
             for (int i = 0; i < unrotatedColCount; i++) { // rows
-                List<TableCell> lastRow = new ArrayList<>();
+                List<Cell> lastRow = new ArrayList<>();
                 for (int j = unrotatedRowCount - 1; j >= 0; j--) { // cols
-                    TableCell cell = cells.get(new TableCellPosition(j, i));
+                    Cell cell = cells.get(new CellPosition(j, i));
                     if (cell != null) {
                         lastRow.add(cell);
                     }
@@ -162,9 +152,9 @@ public class Table extends AbstractTextContainer {
             }
         } else if (rotation == 270) {
             for (int i = unrotatedColCount - 1; i >= 0; i--) { // rows
-                List<TableCell> lastRow = new ArrayList<>();
+                List<Cell> lastRow = new ArrayList<>();
                 for (int j = 0; j < unrotatedRowCount; j++) { // cols
-                    TableCell cell = cells.get(new TableCellPosition(j, i));
+                    Cell cell = cells.get(new CellPosition(j, i));
                     if (cell != null) {
                         lastRow.add(cell);
                     }
@@ -173,9 +163,9 @@ public class Table extends AbstractTextContainer {
             }
         } else {
             for (int i = 0; i < unrotatedRowCount; i++) {
-                List<TableCell> lastRow = new ArrayList<>();
+                List<Cell> lastRow = new ArrayList<>();
                 for (int j = 0; j < unrotatedColCount; j++) {
-                    TableCell cell = cells.get(new TableCellPosition(i, j)); // JAVA_8 use getOrDefault()
+                    Cell cell = cells.get(new CellPosition(i, j)); // JAVA_8 use getOrDefault()
                     if (cell != null) {
                         lastRow.add(cell);
                     }
@@ -189,18 +179,18 @@ public class Table extends AbstractTextContainer {
     }
 
 
-    private void add(TableCell chunk, int row, int col) {
+    private void add(Cell chunk, int row, int col) {
 
         unrotatedRowCount = Math.max(unrotatedRowCount, row + 1);
         unrotatedColCount = Math.max(unrotatedColCount, col + 1);
 
-        TableCellPosition cp = new TableCellPosition(row, col);
+        CellPosition cp = new CellPosition(row, col);
         cells.put(cp, chunk);
 
     }
 
 
-    private void addCells(List<TableCell> cells) {
+    private void addCells(List<Cell> cells) {
 
         if (cells.isEmpty()) {
             return;
@@ -208,7 +198,7 @@ public class Table extends AbstractTextContainer {
 
         cells.removeIf(cell -> cell.getWidth() < 1.1 || cell.getHeight() < 1.1);
 
-        List<List<TableCell>> rowsOfCells = calculateStructure(cells);
+        List<List<Cell>> rowsOfCells = calculateStructure(cells);
 
         for (int i = 0; i < rowsOfCells.size(); i++) {
             for (int j = 0; j < rowsOfCells.get(i).size(); j++) {
@@ -223,11 +213,11 @@ 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<TableCell>> calculateStructure(List<TableCell> cells) {
+    private List<List<Cell>> calculateStructure(List<Cell> cells) {
 
-        List<List<TableCell>> matrix = new ArrayList<>();
+        List<List<Cell>> matrix = new ArrayList<>();
 
         if (cells.isEmpty()) {
             return matrix;
@@ -242,30 +232,30 @@ public class Table extends AbstractTextContainer {
             uniqueY.add(c.getTop());
         });
 
-        var sortedUniqueX = uniqueX.stream().sorted().collect(Collectors.toList());
-        var sortedUniqueY = uniqueY.stream().sorted().collect(Collectors.toList());
+        var sortedUniqueX = uniqueX.stream().sorted().toList();
+        var sortedUniqueY = uniqueY.stream().sorted().toList();
 
         Float prevY = null;
         for (Float y : sortedUniqueY) {
 
-            List<TableCell> row = new ArrayList<>();
+            List<Cell> row = new ArrayList<>();
 
             Float prevX = null;
             for (Float x : sortedUniqueX) {
 
                 if (prevY != null && prevX != null) {
-                    var cell = new TableCell(new Point2D.Float(prevX, prevY), new Point2D.Float(x, y));
+                    var cell = new Cell(new Point2D.Float(prevX, prevY), new Point2D.Float(x, y));
 
                     var intersectionCell = cells.stream().filter(c -> cell.intersects(c) && cell.overlapRatio(c) > 0.1f).findFirst();
-                    if (intersectionCell.isPresent()) {
-                        cell.getTextBlocks().addAll(intersectionCell.get().getTextBlocks());
+                    intersectionCell.ifPresent(value -> cell.getTextBlocks().addAll(value.getTextBlocks()));
+                    if (cell.hasMinimumSize()) {
+                        row.add(cell);
                     }
-                    row.add(cell);
                 }
                 prevX = x;
             }
 
-            if (prevY != null && prevX != null) {
+            if (prevY != null && prevX != null && !row.isEmpty()) {
                 matrix.add(row);
             }
             prevY = y;
@@ -281,22 +271,22 @@ public class Table extends AbstractTextContainer {
     public String getText() {
 
         StringBuilder sb = new StringBuilder();
-        List<List<TableCell>> rows = getRows();
+        List<List<Cell>> rows = getRows();
 
         int i = 0;
-        for (List<TableCell> row : rows) {
+        for (List<Cell> row : rows) {
             if (i != 0) {
                 sb.append("\n");
             }
             if (!row.isEmpty()) {
                 boolean firstColumn = true;
-                for (TableCell column : row) {
+                for (Cell column : row) {
                     if (!firstColumn) {
                         sb.append(",");
                     }
                     if (column != null && column.getTextBlocks() != null) {
                         boolean first = true;
-                        for (ClassificationTextBlock textBlock : column.getTextBlocks()) {
+                        for (TextPageBlock textBlock : column.getTextBlocks()) {
                             if (!first) {
                                 sb.append("\n");
                             }
@@ -317,18 +307,18 @@ public class Table extends AbstractTextContainer {
     public String getTextAsHtml() {
 
         StringBuilder sb = new StringBuilder();
-        List<List<TableCell>> rows = getRows();
+        List<List<Cell>> rows = getRows();
 
         sb.append("<table border=\"1\">");
         int i = 0;
-        for (List<TableCell> row : rows) {
+        for (List<Cell> row : rows) {
             sb.append("\n<tr>");
             if (!row.isEmpty()) {
-                for (TableCell column : row) {
+                for (Cell column : row) {
                     sb.append(i == 0 ? "\n<th>" : "\n<td>");
                     if (column != null && column.getTextBlocks() != null) {
                         boolean first = true;
-                        for (ClassificationTextBlock textBlock : column.getTextBlocks()) {
+                        for (TextPageBlock textBlock : column.getTextBlocks()) {
                             if (!first) {
                                 sb.append("<br />");
                             }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/text/RedTextPosition.java

@@ -1,10 +1,8 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto.text;
+package com.knecon.fforesight.service.layoutparser.processor.classification.model.text;
 
 import org.apache.pdfbox.text.TextPosition;
 import org.springframework.beans.BeanUtils;
 
-import com.dslplatform.json.CompiledJson;
-import com.dslplatform.json.JsonAttribute;
 import com.fasterxml.jackson.annotation.JsonIgnore;
 
 import lombok.AllArgsConstructor;
@@ -17,7 +15,6 @@ import lombok.SneakyThrows;
 @Builder
 @NoArgsConstructor
 @AllArgsConstructor
-@CompiledJson
 public class RedTextPosition {
 
     private String textMatrix;
@@ -39,17 +36,14 @@ public class RedTextPosition {
 
     // not used in reanalysis
     @JsonIgnore
-    @JsonAttribute(ignore = true)
     private float widthOfSpace;
 
     // not used in reanalysis
     @JsonIgnore
-    @JsonAttribute(ignore = true)
     private float fontSizeInPt;
 
     // not used in reanalysis
     @JsonIgnore
-    @JsonAttribute(ignore = true)
     private String fontName;
 
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/text/SearchableText.java

@@ -0,0 +1,48 @@
+package com.knecon.fforesight.service.layoutparser.processor.classification.model.text;
+
+import java.util.ArrayList;
+import java.util.List;
+import com.knecon.fforesight.service.layoutparser.processor.classification.utils.TextNormalizationUtilities;
+
+import lombok.Getter;
+
+@Getter
+public class SearchableText {
+
+    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);
+            sb.append(' ');
+        }
+        String text = sb.toString();
+        text = TextNormalizationUtilities.removeHyphenLineBreaks(text);
+        text = TextNormalizationUtilities.removeLineBreaks(text);
+        text = TextNormalizationUtilities.removeRepeatingWhitespaces(text);
+        return text;
+    }
+
+}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/text/SimplifiedSectionText.java

@@ -0,0 +1,17 @@
+package com.knecon.fforesight.service.layoutparser.processor.classification.model.text;
+
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.Data;
+import lombok.NoArgsConstructor;
+
+@Data
+@Builder
+@NoArgsConstructor
+@AllArgsConstructor
+public class SimplifiedSectionText {
+
+    private int sectionNumber;
+    private String text;
+
+}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/text/SimplifiedText.java

@@ -0,0 +1,20 @@
+package com.knecon.fforesight.service.layoutparser.processor.classification.model.text;
+
+import java.util.ArrayList;
+import java.util.List;
+
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.Data;
+import lombok.NoArgsConstructor;
+
+@Data
+@Builder
+@NoArgsConstructor
+@AllArgsConstructor
+public class SimplifiedText {
+
+    private int numberOfPages;
+    private List<SimplifiedSectionText> sectionTexts = new ArrayList<>();
+
+}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/text/StringFrequencyCounter.java

@@ -1,4 +1,4 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto.text;
+package com.knecon.fforesight.service.layoutparser.processor.classification.model.text;
 
 import java.util.HashMap;
 import java.util.Map;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/text/TextDirection.java

@@ -1,4 +1,4 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto.text;
+package com.knecon.fforesight.service.layoutparser.processor.classification.model.text;
 
 import com.fasterxml.jackson.annotation.JsonCreator;
 import com.fasterxml.jackson.annotation.JsonValue;
@@ -33,13 +33,6 @@ public enum TextDirection {
     }
 
 
-    @com.dslplatform.json.JsonValue
-    public float jsonValue() {
-
-        return getDegrees();
-    }
-
-
     @JsonCreator(mode = JsonCreator.Mode.DELEGATING)
     public static TextDirection fromDegrees(float degrees) {
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/text/TextPageBlock.java

@@ -1,9 +1,16 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto.text;
+package com.knecon.fforesight.service.layoutparser.processor.classification.model.text;
+
+import static java.util.stream.Collectors.toSet;
 
 import java.util.ArrayList;
+import java.util.Comparator;
 import java.util.List;
 
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.AbstractTextContainer;
+import com.fasterxml.jackson.annotation.JsonIgnore;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.AbstractPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.FloatFrequencyCounter;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.PageBlockType;
+import com.knecon.fforesight.service.layoutparser.processor.classification.utils.DoubleComparisons;
 import com.knecon.fforesight.service.layoutparser.processor.classification.utils.TextNormalizationUtilities;
 
 import lombok.AllArgsConstructor;
@@ -13,50 +20,118 @@ import lombok.EqualsAndHashCode;
 import lombok.NoArgsConstructor;
 
 @EqualsAndHashCode(callSuper = true)
+@Data
 @AllArgsConstructor
 @Builder
-@Data
 @NoArgsConstructor
-public class ClassificationTextBlock extends AbstractTextContainer {
+public class TextPageBlock extends AbstractPageBlock {
 
     @Builder.Default
     private List<TextPositionSequence> sequences = new ArrayList<>();
 
+    @JsonIgnore
     private int rotation;
 
-    private int indexOnPage;
-
+    @JsonIgnore
     private String mostPopularWordFont;
 
+    @JsonIgnore
     private String mostPopularWordStyle;
 
+    @JsonIgnore
     private float mostPopularWordFontSize;
 
+    @JsonIgnore
     private float mostPopularWordHeight;
 
+    @JsonIgnore
     private float mostPopularWordSpaceWidth;
 
+    @JsonIgnore
     private float highestFontSize;
 
-    private String classification;
+    @JsonIgnore
+    private PageBlockType classification;
 
 
+    @JsonIgnore
     public TextDirection getDir() {
 
         return sequences.get(0).getDir();
     }
 
+
+    @JsonIgnore
     private float getPageHeight() {
 
         return sequences.get(0).getPageHeight();
     }
 
 
+    @JsonIgnore
     private float getPageWidth() {
 
         return sequences.get(0).getPageWidth();
     }
 
+    public static TextPageBlock merge(List<TextPageBlock> textBlocksToMerge) {
+
+        List<TextPositionSequence> sequences = textBlocksToMerge.stream().map(TextPageBlock::getSequences).flatMap(java.util.Collection::stream).toList();
+        sequences = new ArrayList<>(sequences);
+        return fromTextPositionSequences(sequences);
+    }
+
+    public static TextPageBlock fromTextPositionSequences(List<TextPositionSequence> wordBlockList) {
+
+        TextPageBlock textBlock = null;
+
+        FloatFrequencyCounter lineHeightFrequencyCounter = new FloatFrequencyCounter();
+        FloatFrequencyCounter fontSizeFrequencyCounter = new FloatFrequencyCounter();
+        FloatFrequencyCounter spaceFrequencyCounter = new FloatFrequencyCounter();
+        StringFrequencyCounter fontFrequencyCounter = new StringFrequencyCounter();
+        StringFrequencyCounter styleFrequencyCounter = new StringFrequencyCounter();
+
+        for (TextPositionSequence wordBlock : wordBlockList) {
+
+            lineHeightFrequencyCounter.add(wordBlock.getTextHeight());
+            fontSizeFrequencyCounter.add(wordBlock.getFontSize());
+            spaceFrequencyCounter.add(wordBlock.getSpaceWidth());
+            fontFrequencyCounter.add(wordBlock.getFont());
+            styleFrequencyCounter.add(wordBlock.getFontStyle());
+
+            if (textBlock == null) {
+                textBlock = new TextPageBlock(wordBlock.getMinXDirAdj(),
+                        wordBlock.getMaxXDirAdj(),
+                        wordBlock.getMinYDirAdj(),
+                        wordBlock.getMaxYDirAdj(),
+                        wordBlockList,
+                        wordBlock.getRotation());
+            } else {
+                TextPageBlock spatialEntity = textBlock.union(wordBlock);
+                textBlock.resize(spatialEntity.getMinX(), spatialEntity.getMinY(), spatialEntity.getWidth(), spatialEntity.getHeight());
+            }
+        }
+
+        if (textBlock != null) {
+            textBlock.setMostPopularWordFont(fontFrequencyCounter.getMostPopular());
+            textBlock.setMostPopularWordStyle(styleFrequencyCounter.getMostPopular());
+            textBlock.setMostPopularWordFontSize(fontSizeFrequencyCounter.getMostPopular());
+            textBlock.setMostPopularWordHeight(lineHeightFrequencyCounter.getMostPopular());
+            textBlock.setMostPopularWordSpaceWidth(spaceFrequencyCounter.getMostPopular());
+            textBlock.setHighestFontSize(fontSizeFrequencyCounter.getHighest());
+        }
+
+        if (textBlock != null && textBlock.getSequences() != null && textBlock.getSequences()
+                .stream()
+                .map(t -> DoubleComparisons.round(t.getMinYDirAdj(), 3))
+                .collect(toSet())
+                .size() == 1) {
+            textBlock.getSequences().sort(Comparator.comparing(TextPositionSequence::getMinXDirAdj));
+        }
+        return textBlock;
+    }
+
+
 
     /**
      * Returns the minX value in pdf coordinate system.
@@ -68,6 +143,7 @@ public class ClassificationTextBlock extends AbstractTextContainer {
      *
      * @return the minX value in pdf coordinate system
      */
+    @JsonIgnore
     public float getPdfMinX() {
 
         if (getDir().getDegrees() == 90) {
@@ -83,6 +159,7 @@ public class ClassificationTextBlock 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.
@@ -93,6 +170,7 @@ public class ClassificationTextBlock extends AbstractTextContainer {
      *
      * @return the maxX value in pdf coordinate system
      */
+    @JsonIgnore
     public float getPdfMaxX() {
 
         if (getDir().getDegrees() == 90) {
@@ -118,6 +196,7 @@ public class ClassificationTextBlock extends AbstractTextContainer {
      *
      * @return the minY value in pdf coordinate system
      */
+    @JsonIgnore
     public float getPdfMinY() {
 
         if (getDir().getDegrees() == 90) {
@@ -144,6 +223,7 @@ public class ClassificationTextBlock extends AbstractTextContainer {
      *
      * @return the maxY value in pdf coordinate system
      */
+    @JsonIgnore
     public float getPdfMaxY() {
 
         if (getDir().getDegrees() == 90) {
@@ -159,35 +239,34 @@ public class ClassificationTextBlock extends AbstractTextContainer {
     }
 
 
-    public ClassificationTextBlock(float minX, float maxX, float minY, float maxY, List<TextPositionSequence> sequences, int rotation, int indexOnPage) {
-        super();
-        this.indexOnPage = indexOnPage;
-        super.minX = minX;
-        super.maxX = maxX;
-        super.minY = minY;
-        super.maxY = maxY;
+    public TextPageBlock(float minX, float maxX, float minY, float maxY, List<TextPositionSequence> sequences, int rotation) {
+
+        this.minX = minX;
+        this.maxX = maxX;
+        this.minY = minY;
+        this.maxY = maxY;
         this.sequences = sequences;
         this.rotation = rotation;
     }
 
 
-    public ClassificationTextBlock union(TextPositionSequence r) {
+    public TextPageBlock union(TextPositionSequence r) {
 
-        ClassificationTextBlock union = this.copy();
+        TextPageBlock union = this.copy();
         union.add(r);
         return union;
     }
 
 
-    public ClassificationTextBlock union(ClassificationTextBlock r) {
+    public TextPageBlock union(TextPageBlock r) {
 
-        ClassificationTextBlock union = this.copy();
+        TextPageBlock union = this.copy();
         union.add(r);
         return union;
     }
 
 
-    public void add(ClassificationTextBlock r) {
+    public void add(TextPageBlock r) {
 
         if (r.getMinX() < minX) {
             minX = r.getMinX();
@@ -222,9 +301,9 @@ public class ClassificationTextBlock extends AbstractTextContainer {
     }
 
 
-    public ClassificationTextBlock copy() {
+    public TextPageBlock copy() {
 
-        return new ClassificationTextBlock(minX, maxX, minY, maxY, sequences, rotation, indexOnPage);
+        return new TextPageBlock(minX, maxX, minY, maxY, sequences, rotation);
     }
 
 
@@ -263,6 +342,7 @@ public class ClassificationTextBlock extends AbstractTextContainer {
 
 
     @Override
+    @JsonIgnore
     public String getText() {
 
         StringBuilder sb = new StringBuilder();
@@ -283,4 +363,5 @@ public class ClassificationTextBlock extends AbstractTextContainer {
         return TextNormalizationUtilities.removeHyphenLineBreaks(sb.toString());
 
     }
+
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/text/TextPositionSequence.java

@@ -1,4 +1,4 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto.text;
+package com.knecon.fforesight.service.layoutparser.processor.classification.model.text;
 
 import java.awt.geom.AffineTransform;
 import java.awt.geom.Point2D;
@@ -35,9 +35,16 @@ public class TextPositionSequence implements CharSequence {
     private int rotation;
     private float pageHeight;
     private float pageWidth;
+    private boolean isParagraphStart;
 
 
-    public TextPositionSequence(List<TextPosition> textPositions, int page) {
+    public TextPositionSequence(int page) {
+
+        this.page = page;
+    }
+
+
+    public TextPositionSequence(List<TextPosition> textPositions, int page, boolean isParagraphStart) {
 
         this.textPositions = textPositions.stream().map(RedTextPosition::fromTextPosition).collect(Collectors.toList());
         this.page = page;
@@ -45,6 +52,7 @@ public class TextPositionSequence implements CharSequence {
         this.rotation = textPositions.get(0).getRotation();
         this.pageHeight = textPositions.get(0).getPageHeight();
         this.pageWidth = textPositions.get(0).getPageWidth();
+        this.isParagraphStart = isParagraphStart;
     }
 
 
@@ -64,6 +72,14 @@ public class TextPositionSequence implements CharSequence {
     }
 
 
+    public char charAt(int index, boolean caseInSensitive) {
+
+        RedTextPosition textPosition = textPositionAt(index);
+        String text = textPosition.getUnicode();
+        return caseInSensitive ? text.toLowerCase().charAt(0) : text.charAt(0);
+    }
+
+
     @Override
     public TextPositionSequence subSequence(int start, int end) {
 
@@ -260,7 +276,7 @@ public class TextPositionSequence implements CharSequence {
     @SneakyThrows
     public Rectangle getRectangle() {
 
-        log.debug("ClassificationPage: '{}', Word: '{}', Rotation: '{}', textRotation {}", page, this, rotation, dir);
+        log.debug("Page: '{}', Word: '{}', Rotation: '{}', textRotation {}", page, this, rotation, dir);
 
         float textHeight = getTextHeight();
 
@@ -296,3 +312,4 @@ public class TextPositionSequence implements CharSequence {
     }
 
 }
+

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/model/text/UnclassifiedText.java

@@ -1,4 +1,4 @@
-package com.knecon.fforesight.service.layoutparser.processor.classification.dto.text;
+package com.knecon.fforesight.service.layoutparser.processor.classification.model.text;
 
 import java.util.List;
 
@@ -9,6 +9,6 @@ import lombok.Data;
 @AllArgsConstructor
 public class UnclassifiedText {
 
-    private List<ClassificationTextBlock> textBlocks;
+    private List<TextPageBlock> textBlocks;
 
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/parsing/LegacyPDFStreamEngine.java

@@ -76,7 +76,7 @@ import org.apache.pdfbox.util.Vector;
  * THIS CODE IS DELIBERATELY INCORRECT, USE PDFStreamEngine INSTEAD.
  */
 @SuppressWarnings({"PMD", "checkstyle:all"})
-class LegacyPDFStreamEngine extends PDFStreamEngine {
+public class LegacyPDFStreamEngine extends PDFStreamEngine {
 
     private static final Log LOG = LogFactory.getLog(LegacyPDFStreamEngine.class);
 
@@ -126,7 +126,7 @@ class LegacyPDFStreamEngine extends PDFStreamEngine {
      * This will initialize and process the contents of the stream.
      *
      * @param page the page to process
-     * @throws IOException if there is an error accessing the stream.
+     * @throws java.io.IOException if there is an error accessing the stream.
      */
     @Override
     public void processPage(PDPage page) throws IOException {
@@ -149,7 +149,7 @@ class LegacyPDFStreamEngine extends PDFStreamEngine {
      * written by Ben Litchfield for PDFStreamEngine.
      */
     @Override
-    protected void showGlyph(Matrix textRenderingMatrix, PDFont font, int code,Vector displacement) throws IOException {
+    protected void showGlyph(Matrix textRenderingMatrix, PDFont font, int code, Vector displacement) throws IOException {
         //
         // legacy calculations which were previously in PDFStreamEngine
         //
@@ -165,7 +165,7 @@ class LegacyPDFStreamEngine extends PDFStreamEngine {
 
         float displacementX = displacement.getX();
         // the sorting algorithm is based on the width of the character. As the displacement
-        // for vertical characters doesn't provide any suitable value for it, we have to 
+        // for vertical characters doesn't provide any suitable value for it, we have to
         // calculate our own
         if (font.isVertical()) {
             displacementX = font.getWidth(code) / 1000;
@@ -382,3 +382,4 @@ class LegacyPDFStreamEngine extends PDFStreamEngine {
     }
 
 }
+

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/parsing/PDFAreaTextStripper.java

@@ -1,82 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.processor.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.knecon.fforesight.service.layoutparser.processor.classification.dto.text.TextPositionSequence;
-
-import lombok.Getter;
-import lombok.Setter;
-
-public class PDFAreaTextStripper extends PDFTextStripperByArea {
-
-    @Getter
-    private List<TextPositionSequence> textPositionSequences = new ArrayList<>();
-
-    @Setter
-    private int pageNumber;
-
-
-    public PDFAreaTextStripper() throws IOException {
-
-    }
-
-
-    @Override
-    public void writeString(String text, List<TextPosition> textPositions) throws IOException {
-
-        int startIndex = 0;
-        for (int i = 0; i <= textPositions.size() - 1; i++) {
-
-            if (i == 0 && (textPositions.get(i).getUnicode().equals(" ") || textPositions.get(i).getUnicode().equals("\u00A0"))) {
-                startIndex++;
-                continue;
-            }
-
-            // Strange but sometimes this is happening, for example: Metolachlor2.pdf
-            if (i > 0 && textPositions.get(i).getX() < textPositions.get(i - 1).getX()) {
-                List<TextPosition> sublist = textPositions.subList(startIndex, i);
-                if (!(sublist.isEmpty() || sublist.size() == 1 && (sublist.get(0).getUnicode().equals(" ") || sublist.get(0).getUnicode().equals("\u00A0")))) {
-                    textPositionSequences.add(new TextPositionSequence(sublist, pageNumber));
-                }
-                startIndex = i;
-            }
-
-            if (textPositions.get(i).getRotation() == 0 && i > 0 && textPositions.get(i).getX() > textPositions.get(i - 1).getEndX() + 1) {
-                List<TextPosition> sublist = textPositions.subList(startIndex, i);
-                if (!(sublist.isEmpty() || sublist.size() == 1 && (sublist.get(0).getUnicode().equals(" ") || sublist.get(0).getUnicode().equals("\u00A0")))) {
-                    textPositionSequences.add(new TextPositionSequence(sublist, pageNumber));
-                }
-                startIndex = i;
-            }
-
-            if (i > 0 && (textPositions.get(i).getUnicode().equals(" ") || textPositions.get(i).getUnicode().equals("\u00A0")) && i <= textPositions.size() - 2) {
-                List<TextPosition> sublist = textPositions.subList(startIndex, i);
-                if (!(sublist.isEmpty() || sublist.size() == 1 && (sublist.get(0).getUnicode().equals(" ") || sublist.get(0).getUnicode().equals("\u00A0")))) {
-                    textPositionSequences.add(new TextPositionSequence(sublist, pageNumber));
-                }
-                startIndex = i + 1;
-            }
-        }
-
-        List<TextPosition> sublist = textPositions.subList(startIndex, textPositions.size());
-        if (!sublist.isEmpty() && (sublist.get(sublist.size() - 1).getUnicode().equals(" ") || sublist.get(sublist.size() - 1).getUnicode().equals("\u00A0"))) {
-            sublist = sublist.subList(0, sublist.size() - 1);
-        }
-        if (!(sublist.isEmpty() || sublist.size() == 1 && (sublist.get(0).getUnicode().equals(" ") || sublist.get(0).getUnicode().equals("\u00A0")))) {
-            textPositionSequences.add(new TextPositionSequence(sublist, pageNumber));
-        }
-        super.writeString(text);
-    }
-
-
-    public void clearPositions() {
-
-        textPositionSequences = new ArrayList<>();
-    }
-
-}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/parsing/PDFLinesTextStripper.java

@@ -1,9 +1,9 @@
 package com.knecon.fforesight.service.layoutparser.processor.classification.parsing;
 
+import java.awt.color.CMMException;
 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;
@@ -32,33 +32,30 @@ import org.apache.pdfbox.cos.COSBase;
 import org.apache.pdfbox.cos.COSNumber;
 import org.apache.pdfbox.pdmodel.PDDocument;
 import org.apache.pdfbox.pdmodel.PDPage;
+import org.apache.pdfbox.pdmodel.graphics.color.PDColor;
 import org.apache.pdfbox.text.TextPosition;
 
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.RedTextPosition;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.TextPositionSequence;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.Ruling;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.table.Ruling;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.text.RedTextPosition;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.text.TextPositionSequence;
 
 import lombok.Getter;
 import lombok.Setter;
+import lombok.SneakyThrows;
 import lombok.extern.slf4j.Slf4j;
 
+@Getter
 @Slf4j
 public class PDFLinesTextStripper extends PDFTextStripper {
 
-    @Getter
     private final List<TextPositionSequence> textPositionSequences = new ArrayList<>();
-    @Getter
     private final List<Ruling> rulings = new ArrayList<>();
     private final List<Ruling> graphicsPath = new ArrayList<>();
     @Setter
     protected PDPage pdpage;
-    @Getter
     private int minCharWidth;
-    @Getter
     private int maxCharWidth;
-    @Getter
     private int minCharHeight;
-    @Getter
     private int maxCharHeight;
 
     private float path_x;
@@ -200,8 +197,8 @@ public class PDFLinesTextStripper extends PDFTextStripper {
     private void addVisibleRulings(List<Ruling> path, boolean stroke) throws IOException {
 
         try {
-            if (stroke && !getGraphicsState().getStrokingColor().isPattern() && getGraphicsState().getStrokingColor()
-                    .toRGB() == 0 || !stroke && !getGraphicsState().getNonStrokingColor().isPattern() && getGraphicsState().getNonStrokingColor().toRGB() == 0) {
+            if (stroke && !getGraphicsState().getStrokingColor().isPattern() && isBlack(getGraphicsState().getStrokingColor()) || //
+                    !stroke && !getGraphicsState().getNonStrokingColor().isPattern() && isBlack(getGraphicsState().getNonStrokingColor())) {
                 rulings.addAll(path);
             }
         } catch (UnsupportedOperationException e) {
@@ -212,14 +209,27 @@ public class PDFLinesTextStripper extends PDFTextStripper {
     }
 
 
+    @SneakyThrows
+    private boolean isBlack(PDColor color) {
+
+        try {
+            return color.toRGB() == 0;
+        } catch (CMMException e) {
+            // see https://github.com/haraldk/TwelveMonkeys/issues/124 or https://issues.apache.org/jira/browse/PDFBOX-3531
+            // This is a quick and dirt hack
+            // Happens for file 216.pdf
+            log.debug(e.getMessage());
+            return color.getComponents()[0] == 0 && color.getComponents()[1] == 0 && color.getComponents()[2] == 0 && color.getComponents()[1] == 1;
+        }
+    }
+
+
     @Override
-    public void writeString(String text, List<TextPosition> textPositions) throws IOException {
+    public void writeString(String text, List<TextPosition> textPositions, boolean isParagraphStart) throws IOException {
 
         int startIndex = 0;
         RedTextPosition previous = null;
 
-        textPositions.sort(Comparator.comparing(TextPosition::getXDirAdj));
-
         for (int i = 0; i <= textPositions.size() - 1; i++) {
 
             if (!textPositionSequences.isEmpty()) {
@@ -255,7 +265,7 @@ public class PDFLinesTextStripper extends PDFTextStripper {
                 if (!(sublist.isEmpty() || sublist.size() == 1 && (sublist.get(0).getUnicode().equals(" ") || sublist.get(0).getUnicode().equals("\u00A0") || sublist.get(0)
                         .getUnicode()
                         .equals("\t")))) {
-                    textPositionSequences.add(new TextPositionSequence(sublist, pageNumber));
+                    textPositionSequences.add(new TextPositionSequence(sublist, pageNumber, i == textPositions.size() - 1 && isParagraphStart));
                 }
                 startIndex = i;
             }
@@ -265,7 +275,7 @@ public class PDFLinesTextStripper extends PDFTextStripper {
                 if (!(sublist.isEmpty() || sublist.size() == 1 && (sublist.get(0).getUnicode().equals(" ") || sublist.get(0).getUnicode().equals("\u00A0") || sublist.get(0)
                         .getUnicode()
                         .equals("\t")))) {
-                    textPositionSequences.add(new TextPositionSequence(sublist, pageNumber));
+                    textPositionSequences.add(new TextPositionSequence(sublist, pageNumber, i == textPositions.size() - 1 && isParagraphStart));
                 }
                 startIndex = i;
             }
@@ -281,11 +291,11 @@ 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 textPosition : sublist) {
-                            textPositionSequences.get(textPositionSequences.size() - 1).add(textPosition);
+                        for (TextPosition t : sublist) {
+                            textPositionSequences.get(textPositionSequences.size() - 1).add(t);
                         }
                     } else {
-                        textPositionSequences.add(new TextPositionSequence(sublist, pageNumber));
+                        textPositionSequences.add(new TextPositionSequence(sublist, pageNumber, i == textPositions.size() - 1 && isParagraphStart));
                     }
                 }
                 startIndex = i + 1;
@@ -308,7 +318,7 @@ public class PDFLinesTextStripper extends PDFTextStripper {
                     textPositionSequences.get(textPositionSequences.size() - 1).add(t);
                 }
             } else {
-                textPositionSequences.add(new TextPositionSequence(sublist, pageNumber));
+                textPositionSequences.add(new TextPositionSequence(sublist, pageNumber, isParagraphStart));
             }
         }
         super.writeString(text);
@@ -333,3 +343,4 @@ public class PDFLinesTextStripper extends PDFTextStripper {
     }
 
 }
+

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/parsing/PDFTextStripper.java

@@ -27,6 +27,7 @@ import java.text.Bidi;
 import java.text.Normalizer;
 import java.util.ArrayList;
 import java.util.Collections;
+import java.util.Comparator;
 import java.util.HashMap;
 import java.util.Iterator;
 import java.util.LinkedList;
@@ -240,10 +241,10 @@ public class PDFTextStripper extends LegacyPDFStreamEngine {
         document = doc;
         output = outputStream;
         if (getAddMoreFormatting()) {
-            paragraphEnd = lineSeparator;
+            paragraphEnd = "\n----ParagraphEnd----\n\n";
             pageStart = lineSeparator;
-            articleStart = lineSeparator;
-            articleEnd = lineSeparator;
+            articleStart = "\n----ArticelStart----\n\n";
+            articleEnd = "\n----ArticelEnd----\n\n";
         }
         startDocument(document);
         processPages(document.getPages());
@@ -594,9 +595,14 @@ public class PDFTextStripper extends LegacyPDFStreamEngine {
                     // but this caused a lot of regression test failures. So, I'm leaving it be for
                     // now
                     if (!overlap(positionY, positionHeight, maxYForLine, maxHeightForLine)) {
-                        writeLine(normalize(line));
-                        line.clear();
+                        var normalized = normalize(line);
+//                        normalized.stream().filter(l -> System.out.println(l.getText().contains("Plenarprotokoll 20/24")).findFirst().isPresent()
+
+
                         lastLineStartPosition = handleLineSeparation(current, lastPosition, lastLineStartPosition, maxHeightForLine);
+                        writeLine(normalized, current.isParagraphStart);
+                        line.clear();
+
                         expectedStartOfNextWordX = EXPECTED_START_OF_NEXT_WORD_X_RESET_VALUE;
                         maxYForLine = MAX_Y_FOR_LINE_RESET_VALUE;
                         maxHeightForLine = MAX_HEIGHT_FOR_LINE_RESET_VALUE;
@@ -630,7 +636,24 @@ public class PDFTextStripper extends LegacyPDFStreamEngine {
                     if (startOfPage && lastPosition == null) {
                         writeParagraphStart();// not sure this is correct for RTL?
                     }
+
                     line.add(new LineItem(position));
+
+//                    Collections.sort(line, new Comparator<LineItem>() {
+//
+//                        @Override
+//                        public int compare(LineItem str1, LineItem str2) {
+//                            if(null == str1.getTextPosition()) {
+//                                return 0;
+//                            }
+//                            else if(null == str2.getTextPosition()) {
+//                                return 0;
+//                            }
+//                            return Float.compare(str1.getTextPosition().getX(), str2.getTextPosition().getX());
+//                        }
+//                    });
+
+//                    line.sort(Comparator.comparing(a -> a.getTextPosition() != null && a.getTextPosition().getX()));
                 }
                 maxHeightForLine = Math.max(maxHeightForLine, positionHeight);
                 minYTopForLine = Math.min(minYTopForLine, positionY - positionHeight);
@@ -646,7 +669,7 @@ public class PDFTextStripper extends LegacyPDFStreamEngine {
             }
             // print the final line
             if (line.size() > 0) {
-                writeLine(normalize(line));
+                writeLine(normalize(line), false);
                 writeParagraphEnd();
             }
             endArticle();
@@ -703,7 +726,7 @@ public class PDFTextStripper extends LegacyPDFStreamEngine {
      * @param textPositions The TextPositions belonging to the text.
      * @throws IOException If there is an error when writing the text.
      */
-    protected void writeString(String text, List<TextPosition> textPositions) throws IOException {
+    protected void writeString(String text, List<TextPosition> textPositions, boolean isParagraphEnd) throws IOException {
 
         writeString(text);
     }
@@ -998,7 +1021,7 @@ public class PDFTextStripper extends LegacyPDFStreamEngine {
 
 
     /**
-     * By default the text stripper will attempt to remove text that overlapps each other. Word paints the same
+     * By default, the text stripper will attempt to remove text that overlapps each other. Word paints the same
      * character several times in order to make it look bold. By setting this to false all text will be extracted, which
      * means that certain sections will be duplicated, but better performance will be noticed.
      *
@@ -1385,6 +1408,7 @@ public class PDFTextStripper extends LegacyPDFStreamEngine {
             } else {
                 writeLineSeparator();
                 writeParagraphSeparator();
+                lastLineStartPosition.setEndParagraphWritten();
             }
         } else {
             writeLineSeparator();
@@ -1428,6 +1452,10 @@ public class PDFTextStripper extends LegacyPDFStreamEngine {
             float newXVal = multiplyFloat(getIndentThreshold(), position.getTextPosition().getWidthOfSpace());
             float positionWidth = multiplyFloat(0.25f, position.getTextPosition().getWidth());
 
+//            if(xGap < 0){
+//                result = true;
+//            }
+//            else
             if (yGap > newYVal) {
                 result = true;
             } else if (xGap > newXVal) {
@@ -1636,12 +1664,13 @@ public class PDFTextStripper extends LegacyPDFStreamEngine {
      * @param line a list with the words of the given line
      * @throws IOException if something went wrong
      */
-    private void writeLine(List<WordWithTextPositions> line) throws IOException {
+    private void writeLine(List<WordWithTextPositions> line, boolean isParagraphEnd) throws IOException {
 
         int numberOfStrings = line.size();
         for (int i = 0; i < numberOfStrings; i++) {
             WordWithTextPositions word = line.get(i);
-            writeString(word.getText(), word.getTextPositions());
+            word.getTextPositions().sort(Comparator.comparing(TextPosition::getX));
+            writeString(word.getText(), word.getTextPositions(), isParagraphEnd && i == numberOfStrings - 1);
             if (i < numberOfStrings - 1) {
                 writeWordSeparator();
             }
@@ -1963,6 +1992,8 @@ public class PDFTextStripper extends LegacyPDFStreamEngine {
         private boolean isHangingIndent = false;
         private boolean isArticleStart = false;
 
+        private boolean endParagraphWritten = false;
+
         private TextPosition position = null;
 
 
@@ -2024,6 +2055,16 @@ public class PDFTextStripper extends LegacyPDFStreamEngine {
         }
 
 
+        public boolean isEndParagraphWritten() {
+
+            return endParagraphWritten;
+        }
+
+        public void setEndParagraphWritten(){
+            endParagraphWritten = true;
+        }
+
+
         /**
          * Sets the isArticleStart() flag to true.
          */
@@ -2065,3 +2106,4 @@ public class PDFTextStripper extends LegacyPDFStreamEngine {
     }
 
 }
+

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/service/BlockificationService.java

@@ -1,29 +1,29 @@
 package com.knecon.fforesight.service.layoutparser.processor.classification.service;
 
-import static java.util.stream.Collectors.toSet;
-
 import java.util.ArrayList;
-import java.util.Comparator;
 import java.util.Iterator;
+import java.util.LinkedList;
 import java.util.List;
+import java.util.regex.Matcher;
+import java.util.regex.Pattern;
 
 import org.springframework.stereotype.Service;
 
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.ClassificationPage;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.ClassificationTextBlock;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.FloatFrequencyCounter;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.TextBlockOrientation;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.StringFrequencyCounter;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.TextPositionSequence;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.AbstractTextContainer;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.Ruling;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.AbstractPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.ClassificationPage;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.Orientation;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.table.Ruling;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.text.TextPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.text.TextPositionSequence;
 import com.knecon.fforesight.service.layoutparser.processor.classification.utils.RulingTextDirAdjustUtil;
 
 @Service
 @SuppressWarnings("all")
 public class BlockificationService {
 
-    static final float THRESHOLD = 1f;
+    private static final float THRESHOLD = 1f;
+    private static final float Y_GAP_SPLIT_HEIGHT_MODIFIER = 1.25f;
+    private static final int X_GAP_SPLIT_CONSTANT = 50;
 
 
     /**
@@ -38,48 +38,150 @@ public class BlockificationService {
      */
     public ClassificationPage blockify(List<TextPositionSequence> textPositions, List<Ruling> horizontalRulingLines, List<Ruling> verticalRulingLines) {
 
+        List<TextPageBlock> classificationTextBlocks = constructFineGranularTextPageBlocks(textPositions, horizontalRulingLines, verticalRulingLines);
+
+        classificationTextBlocks = mergeFineGranularTextPageBlocks(classificationTextBlocks);
+
+        return new ClassificationPage(new ArrayList<>(classificationTextBlocks.stream().map(classificationTextBlock -> (AbstractPageBlock) classificationTextBlock).toList()));
+    }
+
+
+    private List<TextPageBlock> mergeFineGranularTextPageBlocks(List<TextPageBlock> classificationTextBlocks) {
+
+        if (classificationTextBlocks.isEmpty()) {
+            return new ArrayList<>();
+        }
+        List<List<TextPageBlock>> textBlocksToMerge = new LinkedList<>();
+        List<TextPageBlock> currentTextBlocksToMerge = new LinkedList<>();
+        textBlocksToMerge.add(currentTextBlocksToMerge);
+        TextPageBlock previousTextBlock = null;
+        for (TextPageBlock currentTextBlock : classificationTextBlocks) {
+            if (previousTextBlock == null) {
+                currentTextBlocksToMerge.add(currentTextBlock);
+                previousTextBlock = currentTextBlock;
+                continue;
+            }
+            boolean alignsXRight = Math.abs(currentTextBlock.getPdfMaxX() - previousTextBlock.getPdfMaxX()) < 1;
+            boolean smallYGap = Math.abs(currentTextBlock.getPdfMaxY() - previousTextBlock.getPdfMinY()) < 5;
+            if (alignsXRight && smallYGap) {
+                currentTextBlocksToMerge.add(currentTextBlock);
+            } else {
+                currentTextBlocksToMerge = new LinkedList<>();
+                currentTextBlocksToMerge.add(currentTextBlock);
+                textBlocksToMerge.add(currentTextBlocksToMerge);
+            }
+            previousTextBlock = currentTextBlock;
+        }
+        return textBlocksToMerge.stream().map(TextPageBlock::merge).toList();
+    }
+
+
+    private void assignOrientations(List<TextPageBlock> classificationTextBlocks) {
+
+        Iterator<TextPageBlock> itty = classificationTextBlocks.iterator();
+
+        TextPageBlock previousLeft = null;
+        TextPageBlock previousRight = null;
+        while (itty.hasNext()) {
+            TextPageBlock block = (TextPageBlock) itty.next();
+
+            if (previousLeft != null && block.getOrientation().equals(Orientation.LEFT)) {
+                if (previousLeft.getMinY() > block.getMinY() && block.getMaxY() + block.getMostPopularWordHeight() > previousLeft.getMinY()) {
+                    previousLeft.add(block);
+                    itty.remove();
+                    continue;
+                }
+            }
+
+            if (previousRight != null && block.getOrientation().equals(Orientation.RIGHT)) {
+                if (previousRight.getMinY() > block.getMinY() && block.getMaxY() + block.getMostPopularWordHeight() > previousRight.getMinY()) {
+                    previousRight.add(block);
+                    itty.remove();
+                    continue;
+                }
+            }
+
+            if (block.getOrientation().equals(Orientation.LEFT)) {
+                previousLeft = block;
+            } else if (block.getOrientation().equals(Orientation.RIGHT)) {
+                previousRight = block;
+            }
+        }
+
+        itty = classificationTextBlocks.iterator();
+        TextPageBlock previous = null;
+        while (itty.hasNext()) {
+            TextPageBlock block = (TextPageBlock) 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()
+                    .equals(Orientation.RIGHT) && equalsWithThreshold(block.getMaxY(), previous.getMaxY())) {
+                previous.add(block);
+                itty.remove();
+                continue;
+            }
+
+            previous = block;
+        }
+    }
+
+
+    private List<TextPageBlock> constructFineGranularTextPageBlocks(List<TextPositionSequence> textPositions,
+                                                                                        List<Ruling> horizontalRulingLines,
+                                                                                        List<Ruling> verticalRulingLines) {
+
         int indexOnPage = 0;
-        List<TextPositionSequence> chunkWords = new ArrayList<>();
-        List<AbstractTextContainer> chunkBlockList1 = new ArrayList<>();
+        List<TextPositionSequence> wordClusterToCombine = new ArrayList<>();
+        List<TextPageBlock> classificationTextBlocks = new ArrayList<>();
 
         float minX = 1000, maxX = 0, minY = 1000, maxY = 0;
         TextPositionSequence prev = null;
 
+        var listIdentitifier = Pattern.compile("\\b(?:[1-9]|1\\d|20|[ivxlc]|[a-z])\\s*(?:[.)])", Pattern.CASE_INSENSITIVE);
+
         boolean wasSplitted = false;
         Float splitX1 = null;
         for (TextPositionSequence word : textPositions) {
 
-            boolean lineSeparation = word.getMinYDirAdj() - maxY > word.getHeight() * 1.25;
+            Matcher listIdentifierPattern = listIdentitifier.matcher(word.toString());
+
+            boolean yGap = word.getMinYDirAdj() - maxY > word.getHeight() * Y_GAP_SPLIT_HEIGHT_MODIFIER;
+            boolean sameLine = prev != null && equalsWithThreshold(prev.getMinYDirAdj(), word.getMinYDirAdj());
+            boolean positiveXGapInline = prev != null && maxX + X_GAP_SPLIT_CONSTANT < word.getMinXDirAdj() && sameLine;
+            boolean negativeXGap = prev != null && word.getMinXDirAdj() - minX < -5;
             boolean startFromTop = prev != null && word.getMinYDirAdj() < prev.getMinYDirAdj() - prev.getTextHeight();
-            boolean splitByX = prev != null && maxX + 50 < word.getMinXDirAdj() && prev.getMinYDirAdj() == word.getMinYDirAdj();
-            boolean xIsBeforeFirstX = prev != null && word.getMinXDirAdj() < minX;
             boolean newLineAfterSplit = prev != null && word.getMinYDirAdj() != prev.getMinYDirAdj() && wasSplitted && splitX1 != word.getMinXDirAdj();
-            boolean isSplitByRuling = isSplitByRuling(minX, minY, maxX, maxY, word, horizontalRulingLines, verticalRulingLines);
+            boolean splitByRuling = isSplitByRuling(minX, minY, maxX, maxY, word, horizontalRulingLines, verticalRulingLines);
             boolean splitByDir = prev != null && !prev.getDir().equals(word.getDir());
+            boolean fontChange = prev != null && (!word.getFont().equals(prev.getFont()) || !word.getFontStyle()
+                    .equals(prev.getFontStyle()) || word.getFontSize() != prev.getFontSize());
+            boolean newline = prev != null && Math.abs(word.getMinYDirAdj() - prev.getMinYDirAdj()) > word.getHeight();
+            boolean isListIdentifier = listIdentifierPattern.matches();
 
-            if (prev != null && (lineSeparation || startFromTop || splitByX || splitByDir || isSplitByRuling)) {
+            if (prev != null && (prev.isParagraphStart() || negativeXGap || positiveXGapInline || yGap || startFromTop || splitByRuling || (newline && (fontChange || isListIdentifier)))) {
+//            if (prev != null && (lineSeparation || startFromTop || splitByX || splitByDir || isSplitByRuling)) {
 
-                TextBlockOrientation prevOrientation = null;
-                if (!chunkBlockList1.isEmpty()) {
-                    prevOrientation = chunkBlockList1.get(chunkBlockList1.size() - 1).getOrientation();
+                Orientation prevOrientation = null;
+                if (!classificationTextBlocks.isEmpty()) {
+                    prevOrientation = classificationTextBlocks.get(classificationTextBlocks.size() - 1).getOrientation();
                 }
 
-                ClassificationTextBlock cb1 = buildTextBlock(chunkWords, indexOnPage);
-                indexOnPage++;
+                TextPageBlock classificationTextBlock = TextPageBlock.fromTextPositionSequences(wordClusterToCombine);
 
-                chunkBlockList1.add(cb1);
-                chunkWords = new ArrayList<>();
+                classificationTextBlocks.add(classificationTextBlock);
+                wordClusterToCombine = new ArrayList<>();
 
-                if (splitByX && !isSplitByRuling) {
+                if (positiveXGapInline && !splitByRuling) {
                     wasSplitted = true;
-                    cb1.setOrientation(TextBlockOrientation.LEFT);
+                    classificationTextBlock.setOrientation(Orientation.LEFT);
                     splitX1 = word.getMinXDirAdj();
-                } else if (newLineAfterSplit && !isSplitByRuling) {
+                } else if (newLineAfterSplit && !splitByRuling) {
                     wasSplitted = false;
-                    cb1.setOrientation(TextBlockOrientation.RIGHT);
+                    classificationTextBlock.setOrientation(Orientation.RIGHT);
                     splitX1 = null;
-                } else if (prevOrientation != null && prevOrientation.equals(TextBlockOrientation.RIGHT) && (lineSeparation || !startFromTop || !splitByX || !newLineAfterSplit || !isSplitByRuling)) {
-                    cb1.setOrientation(TextBlockOrientation.LEFT);
+                } else if (prevOrientation != null && prevOrientation.equals(Orientation.RIGHT) && (yGap || !startFromTop || !positiveXGapInline || !newLineAfterSplit || !splitByRuling)) {
+                    classificationTextBlock.setOrientation(Orientation.LEFT);
                 }
 
                 minX = 1000;
@@ -89,7 +191,7 @@ public class BlockificationService {
                 prev = null;
             }
 
-            chunkWords.add(word);
+            wordClusterToCombine.add(word);
 
             prev = word;
             if (word.getMinXDirAdj() < minX) {
@@ -106,58 +208,11 @@ public class BlockificationService {
             }
         }
 
-        ClassificationTextBlock cb1 = buildTextBlock(chunkWords, indexOnPage);
-        if (cb1 != null) {
-            chunkBlockList1.add(cb1);
+        TextPageBlock classificationTextBlock = TextPageBlock.fromTextPositionSequences(wordClusterToCombine);
+        if (classificationTextBlock != null) {
+            classificationTextBlocks.add(classificationTextBlock);
         }
-
-        Iterator<AbstractTextContainer> itty = chunkBlockList1.iterator();
-
-        ClassificationTextBlock previousLeft = null;
-        ClassificationTextBlock previousRight = null;
-        while (itty.hasNext()) {
-            ClassificationTextBlock block = (ClassificationTextBlock) itty.next();
-
-            if (previousLeft != null && block.getOrientation().equals(TextBlockOrientation.LEFT)) {
-                if (previousLeft.getMinY() > block.getMinY() && block.getMaxY() + block.getMostPopularWordHeight() > previousLeft.getMinY()) {
-                    previousLeft.add(block);
-                    itty.remove();
-                    continue;
-                }
-            }
-
-            if (previousRight != null && block.getOrientation().equals(TextBlockOrientation.RIGHT)) {
-                if (previousRight.getMinY() > block.getMinY() && block.getMaxY() + block.getMostPopularWordHeight() > previousRight.getMinY()) {
-                    previousRight.add(block);
-                    itty.remove();
-                    continue;
-                }
-            }
-
-            if (block.getOrientation().equals(TextBlockOrientation.LEFT)) {
-                previousLeft = block;
-            } else if (block.getOrientation().equals(TextBlockOrientation.RIGHT)) {
-                previousRight = block;
-            }
-        }
-
-        itty = chunkBlockList1.iterator();
-        ClassificationTextBlock previous = null;
-        while (itty.hasNext()) {
-            ClassificationTextBlock block = (ClassificationTextBlock) itty.next();
-
-            if (previous != null && previous.getOrientation().equals(TextBlockOrientation.LEFT) && block.getOrientation().equals(TextBlockOrientation.LEFT) && equalsWithThreshold(block.getMaxY(),
-                    previous.getMaxY()) || previous != null && previous.getOrientation().equals(TextBlockOrientation.LEFT) && block.getOrientation()
-                    .equals(TextBlockOrientation.RIGHT) && equalsWithThreshold(block.getMaxY(), previous.getMaxY())) {
-                previous.add(block);
-                itty.remove();
-                continue;
-            }
-
-            previous = block;
-        }
-
-        return new ClassificationPage(chunkBlockList1);
+        return classificationTextBlocks;
     }
 
 
@@ -167,54 +222,6 @@ public class BlockificationService {
     }
 
 
-    private ClassificationTextBlock buildTextBlock(List<TextPositionSequence> wordBlockList, int indexOnPage) {
-
-        ClassificationTextBlock textBlock = null;
-
-        FloatFrequencyCounter lineHeightFrequencyCounter = new FloatFrequencyCounter();
-        FloatFrequencyCounter fontSizeFrequencyCounter = new FloatFrequencyCounter();
-        FloatFrequencyCounter spaceFrequencyCounter = new FloatFrequencyCounter();
-        StringFrequencyCounter fontFrequencyCounter = new StringFrequencyCounter();
-        StringFrequencyCounter styleFrequencyCounter = new StringFrequencyCounter();
-
-        for (TextPositionSequence wordBlock : wordBlockList) {
-
-            lineHeightFrequencyCounter.add(wordBlock.getTextHeight());
-            fontSizeFrequencyCounter.add(wordBlock.getFontSize());
-            spaceFrequencyCounter.add(wordBlock.getSpaceWidth());
-            fontFrequencyCounter.add(wordBlock.getFont());
-            styleFrequencyCounter.add(wordBlock.getFontStyle());
-
-            if (textBlock == null) {
-                textBlock = new ClassificationTextBlock(wordBlock.getMinXDirAdj(),
-                        wordBlock.getMaxXDirAdj(),
-                        wordBlock.getMinYDirAdj(),
-                        wordBlock.getMaxYDirAdj(),
-                        wordBlockList,
-                        wordBlock.getRotation(),
-                        indexOnPage);
-            } else {
-                ClassificationTextBlock spatialEntity = textBlock.union(wordBlock);
-                textBlock.resize(spatialEntity.getMinX(), spatialEntity.getMinY(), spatialEntity.getWidth(), spatialEntity.getHeight());
-            }
-        }
-
-        if (textBlock != null) {
-            textBlock.setMostPopularWordFont(fontFrequencyCounter.getMostPopular());
-            textBlock.setMostPopularWordStyle(styleFrequencyCounter.getMostPopular());
-            textBlock.setMostPopularWordFontSize(fontSizeFrequencyCounter.getMostPopular());
-            textBlock.setMostPopularWordHeight(lineHeightFrequencyCounter.getMostPopular());
-            textBlock.setMostPopularWordSpaceWidth(spaceFrequencyCounter.getMostPopular());
-            textBlock.setHighestFontSize(fontSizeFrequencyCounter.getHighest());
-        }
-
-        if (textBlock != null && textBlock.getSequences() != null && textBlock.getSequences().stream().map(t -> round(t.getMinYDirAdj(), 3)).collect(toSet()).size() == 1) {
-            textBlock.getSequences().sort(Comparator.comparing(TextPositionSequence::getMinXDirAdj));
-        }
-        return textBlock;
-    }
-
-
     private boolean isSplitByRuling(float minX,
                                     float minY,
                                     float maxX,
@@ -269,11 +276,5 @@ public class BlockificationService {
         return false;
     }
 
-
-    private double round(float value, int decimalPoints) {
-
-        var d = Math.pow(10, decimalPoints);
-        return Math.round(value * d) / d;
-    }
-
 }
+

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/service/BodyTextFrameService.java

@@ -1,52 +1,56 @@
 package com.knecon.fforesight.service.layoutparser.processor.classification.service;
+
 import java.util.List;
 
 import org.springframework.stereotype.Service;
 
-import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.Rectangle;
 import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.Point;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.ClassificationPage;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.FloatFrequencyCounter;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.ClassificationTextBlock;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.AbstractTextContainer;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.TableCell;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.Table;
+import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.Rectangle;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.AbstractPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.ClassificationPage;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.FloatFrequencyCounter;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.table.Cell;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.table.TablePageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.text.TextPageBlock;
 import com.knecon.fforesight.service.layoutparser.processor.classification.utils.PositionUtils;
 
 @Service
 public class BodyTextFrameService {
 
+    private static final float APPROXIMATE_HEADER_LINE_COUNT = 2.0f;
+
+
     /**
-     * Adjusts and sets the body text frame to a classificationPage.
-     * Note: This needs to use Pdf Coordinate System where {0,0} rotated with the classificationPage rotation.
+     * Adjusts and sets the body text frame to a page.
+     * Note: This needs to use Pdf Coordinate System where {0,0} rotated with the page rotation.
      * 0 -> LowerLeft
      * 90 -> UpperLeft
      * 180 -> UpperRight
      * 270 -> LowerRight
-     * The aspect ratio of the classificationPage is also regarded.
+     * The aspect ratio of the page is also regarded.
      *
-     * @param classificationPage                   The classificationPage
+     * @param page                   The page
      * @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(ClassificationPage classificationPage, Rectangle bodyTextFrame, Rectangle landscapeBodyTextFrame) {
+    public void setBodyTextFrameAdjustedToPage(ClassificationPage page, Rectangle bodyTextFrame, Rectangle landscapeBodyTextFrame) {
 
-        Rectangle textFrame = classificationPage.isLandscape() ? landscapeBodyTextFrame : bodyTextFrame;
+        Rectangle textFrame = page.isLandscape() ? landscapeBodyTextFrame : bodyTextFrame;
 
-        if (classificationPage.getPageWidth() > classificationPage.getPageHeight() && classificationPage.getRotation() == 270) {
-            textFrame = new Rectangle(new Point(textFrame.getTopLeft().getY(), classificationPage.getPageHeight() - textFrame.getTopLeft().getX() - textFrame.getWidth()),
+        if (page.getPageWidth() > page.getPageHeight() && page.getRotation() == 270) {
+            textFrame = new Rectangle(new Point(textFrame.getTopLeft().getY(), page.getPageHeight() - textFrame.getTopLeft().getX() - textFrame.getWidth()),
                     textFrame.getHeight(),
                     textFrame.getWidth(),
                     0);
-        } else if (classificationPage.getPageWidth() > classificationPage.getPageHeight() && classificationPage.getRotation() != 0) {
-            textFrame = new Rectangle(new Point(textFrame.getTopLeft().getY(), textFrame.getTopLeft().getX()), textFrame.getHeight(), textFrame.getWidth(), classificationPage.getPageNumber());
-        } else if (classificationPage.getRotation() == 180) {
-            textFrame = new Rectangle(new Point(textFrame.getTopLeft().getX(), classificationPage.getPageHeight() - textFrame.getTopLeft().getY() - textFrame.getHeight()),
+        } else if (page.getPageWidth() > page.getPageHeight() && page.getRotation() != 0) {
+            textFrame = new Rectangle(new Point(textFrame.getTopLeft().getY(), textFrame.getTopLeft().getX()), textFrame.getHeight(), textFrame.getWidth(), page.getPageNumber());
+        } else if (page.getRotation() == 180) {
+            textFrame = new Rectangle(new Point(textFrame.getTopLeft().getX(), page.getPageHeight() - textFrame.getTopLeft().getY() - textFrame.getHeight()),
                     textFrame.getWidth(),
                     textFrame.getHeight(),
                     0);
         }
-        classificationPage.setBodyTextFrame(textFrame);
+        page.setBodyTextFrame(textFrame);
     }
 
 
@@ -59,50 +63,50 @@ public class BodyTextFrameService {
      * 270 -> LowerRight
      * The aspect ratio of the page is also regarded.
      *
-     * @param classificationPages                   List of all classificationPages
+     * @param pages                   List of all pages
      * @param documentFontSizeCounter Statistics of the document
      * @param landscape               Calculate for landscape or portrait
      * @return Rectangle of the text frame
      */
-    public Rectangle calculateBodyTextFrame(List<ClassificationPage> classificationPages, FloatFrequencyCounter documentFontSizeCounter, boolean landscape) {
+    public Rectangle calculateBodyTextFrame(List<ClassificationPage> pages, FloatFrequencyCounter documentFontSizeCounter, boolean landscape) {
 
         BodyTextFrameExpansionsRectangle expansionsRectangle = new BodyTextFrameExpansionsRectangle();
 
-        for (ClassificationPage classificationPage : classificationPages) {
+        for (ClassificationPage page : pages) {
 
-            if (classificationPage.getTextBlocks().isEmpty() || landscape != classificationPage.isLandscape()) {
+            if (page.getTextBlocks().isEmpty() || landscape != page.isLandscape()) {
                 continue;
             }
 
-            for (AbstractTextContainer container : classificationPage.getTextBlocks()) {
+            for (AbstractPageBlock container : page.getTextBlocks()) {
 
-                if (container instanceof ClassificationTextBlock) {
-                    ClassificationTextBlock textBlock = (ClassificationTextBlock) container;
+                if (container instanceof TextPageBlock) {
+                    TextPageBlock textBlock = (TextPageBlock) container;
                     if (textBlock.getMostPopularWordFont() == null || textBlock.getMostPopularWordStyle() == null) {
                         continue;
                     }
 
                     float approxLineCount = PositionUtils.getApproxLineCount(textBlock);
-                    if (approxLineCount < 2.9f) {
+                    if (approxLineCount < APPROXIMATE_HEADER_LINE_COUNT) {
                         continue;
                     }
 
                     if (documentFontSizeCounter.getMostPopular() != null && textBlock.getMostPopularWordFontSize() >= documentFontSizeCounter.getMostPopular()) {
 
-                        expandRectangle(textBlock, classificationPage, expansionsRectangle);
+                        expandRectangle(textBlock, page, expansionsRectangle);
                     }
                 }
 
-                if (container instanceof Table) {
-                    Table table = (Table) container;
-                    for (List<TableCell> row : table.getRows()) {
-                        for (TableCell cell : row) {
+                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 (ClassificationTextBlock textBlock : cell.getTextBlocks()) {
-                                expandRectangle(textBlock, classificationPage, expansionsRectangle);
+                            for (TextPageBlock textBlock : cell.getTextBlocks()) {
+                                expandRectangle(textBlock, page, expansionsRectangle);
                             }
                         }
                     }
@@ -116,9 +120,9 @@ public class BodyTextFrameService {
     }
 
 
-    private void expandRectangle(ClassificationTextBlock textBlock, ClassificationPage classificationPage, BodyTextFrameExpansionsRectangle expansionsRectangle) {
+    private void expandRectangle(TextPageBlock textBlock, ClassificationPage page, BodyTextFrameExpansionsRectangle expansionsRectangle) {
 
-        if (classificationPage.getPageWidth() > classificationPage.getPageHeight() && classificationPage.getRotation() != 0) {
+        if (page.getPageWidth() > page.getPageHeight() && page.getRotation() != 0) {
             if (textBlock.getPdfMinY() < expansionsRectangle.minX) {
                 expansionsRectangle.minX = textBlock.getPdfMinY();
             }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/service/ClassificationService.java

@@ -6,10 +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.knecon.fforesight.service.layoutparser.processor.classification.dto.ClassificationDocument;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.ClassificationPage;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.ClassificationTextBlock;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.AbstractTextContainer;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.AbstractPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.ClassificationDocument;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.ClassificationPage;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.PageBlockType;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.text.TextPageBlock;
 import com.knecon.fforesight.service.layoutparser.processor.classification.utils.PositionUtils;
 
 import lombok.RequiredArgsConstructor;
@@ -31,43 +32,40 @@ public class ClassificationService {
 
         log.debug("Document FontSize counters are: {}", document.getFontSizeCounter().getCountPerValue());
 
-        for (ClassificationPage classificationPage : document.getPages()) {
-            bodyTextFrameService.setBodyTextFrameAdjustedToPage(classificationPage, bodyTextFrame, landscapeBodyTextFrame);
-            classifyPage(classificationPage, document, headlineFontSizes);
+        for (ClassificationPage page : document.getPages()) {
+            bodyTextFrameService.setBodyTextFrameAdjustedToPage(page, bodyTextFrame, landscapeBodyTextFrame);
+            classifyPage(page, document, headlineFontSizes);
         }
     }
 
 
-    public void classifyPage(ClassificationPage classificationPage, ClassificationDocument document, List<Float> headlineFontSizes) {
+    public void classifyPage(ClassificationPage page, ClassificationDocument document, List<Float> headlineFontSizes) {
 
-        for (AbstractTextContainer textBlock : classificationPage.getTextBlocks()) {
-            if (textBlock instanceof ClassificationTextBlock) {
-                classifyBlock((ClassificationTextBlock) textBlock, classificationPage, document, headlineFontSizes);
+        for (AbstractPageBlock textBlock : page.getTextBlocks()) {
+            if (textBlock instanceof TextPageBlock) {
+                classifyBlock((TextPageBlock) textBlock, page, document, headlineFontSizes);
             }
         }
     }
 
 
-    public void classifyBlock(ClassificationTextBlock textBlock, ClassificationPage page, ClassificationDocument document, List<Float> headlineFontSizes) {
+    public void classifyBlock(TextPageBlock textBlock, ClassificationPage page, ClassificationDocument document, List<Float> headlineFontSizes) {
 
         var bodyTextFrame = page.getBodyTextFrame();
 
         if (document.getFontSizeCounter().getMostPopular() == null) {
-            textBlock.setClassification("Other");
+            textBlock.setClassification(PageBlockType.OTHER);
             return;
         }
-        if (PositionUtils.isOverBodyTextFrame(bodyTextFrame, textBlock, page.getRotation()) && (document.getFontSizeCounter()
-                .getMostPopular() == null || textBlock.getHighestFontSize() <= document.getFontSizeCounter().getMostPopular())) {
-            textBlock.setClassification("Header");
-
-        } else if (PositionUtils.isUnderBodyTextFrame(bodyTextFrame, textBlock, page.getRotation()) && (document.getFontSizeCounter()
-                .getMostPopular() == null || textBlock.getHighestFontSize() <= document.getFontSizeCounter().getMostPopular())) {
-            textBlock.setClassification("Footer");
+        if (PositionUtils.isOverBodyTextFrame(bodyTextFrame, textBlock, page.getRotation())) {
+            textBlock.setClassification(PageBlockType.HEADER);
+        } else if (PositionUtils.isUnderBodyTextFrame(bodyTextFrame, textBlock, page.getRotation())) {
+            textBlock.setClassification(PageBlockType.FOOTER);
         } else if (page.getPageNumber() == 1 && (PositionUtils.getHeightDifferenceBetweenChunkWordAndDocumentWord(textBlock,
                 document.getTextHeightCounter().getMostPopular()) > 2.5 && textBlock.getHighestFontSize() > document.getFontSizeCounter().getMostPopular() || page.getTextBlocks()
                 .size() == 1)) {
             if (!Pattern.matches("[0-9]+", textBlock.toString())) {
-                textBlock.setClassification("Title");
+                textBlock.setClassification(PageBlockType.TITLE);
             }
         } else if (textBlock.getMostPopularWordFontSize() > document.getFontSizeCounter()
                 .getMostPopular() && PositionUtils.getApproxLineCount(textBlock) < 4.9 && (textBlock.getMostPopularWordStyle().equals("bold") || !document.getFontStyleCounter()
@@ -80,36 +78,34 @@ public class ClassificationService {
 
             for (int i = 1; i <= headlineFontSizes.size(); i++) {
                 if (textBlock.getMostPopularWordFontSize() == headlineFontSizes.get(i - 1)) {
-                    textBlock.setClassification("H " + i);
+                    textBlock.setClassification(PageBlockType.getHeadlineType(i));
                     document.setHeadlines(true);
                 }
             }
-        } else if (!textBlock.getText().startsWith("Table ") && !textBlock.getText().startsWith("Figure ") && PositionUtils.isWithinBodyTextFrame(bodyTextFrame,
-                textBlock) && textBlock.getMostPopularWordStyle().equals("bold") && !document.getFontStyleCounter()
-                .getMostPopular()
-                .equals("bold") && PositionUtils.getApproxLineCount(textBlock) < 2.9 && textBlock.getSequences()
+        } else if (!textBlock.getText().startsWith("Figure ") && PositionUtils.isWithinBodyTextFrame(bodyTextFrame, textBlock) && textBlock.getMostPopularWordStyle()
+                .equals("bold") && !document.getFontStyleCounter().getMostPopular().equals("bold") && PositionUtils.getApproxLineCount(textBlock) < 2.9 && textBlock.getSequences()
                 .get(0)
                 .getTextPositions()
                 .get(0)
                 .getFontSizeInPt() >= textBlock.getMostPopularWordFontSize()) {
-            textBlock.setClassification("H " + (headlineFontSizes.size() + 1));
+            textBlock.setClassification(PageBlockType.getHeadlineType(headlineFontSizes.size() + 1));
             document.setHeadlines(true);
         } else if (PositionUtils.isWithinBodyTextFrame(bodyTextFrame, textBlock) && textBlock.getMostPopularWordFontSize() == document.getFontSizeCounter()
                 .getMostPopular() && textBlock.getMostPopularWordStyle().equals("bold") && !document.getFontStyleCounter().getMostPopular().equals("bold")) {
-            textBlock.setClassification("TextBlock Bold");
+            textBlock.setClassification(PageBlockType.PARAGRAPH_BOLD);
         } else if (PositionUtils.isWithinBodyTextFrame(bodyTextFrame, textBlock) && textBlock.getMostPopularWordFont()
                 .equals(document.getFontCounter().getMostPopular()) && textBlock.getMostPopularWordStyle()
                 .equals(document.getFontStyleCounter().getMostPopular()) && textBlock.getMostPopularWordFontSize() == document.getFontSizeCounter().getMostPopular()) {
-            textBlock.setClassification("TextBlock");
+            textBlock.setClassification(PageBlockType.PARAGRAPH);
         } else if (PositionUtils.isWithinBodyTextFrame(bodyTextFrame, textBlock) && textBlock.getMostPopularWordFontSize() == document.getFontSizeCounter()
                 .getMostPopular() && textBlock.getMostPopularWordStyle().equals("italic") && !document.getFontStyleCounter()
                 .getMostPopular()
                 .equals("italic") && PositionUtils.getApproxLineCount(textBlock) < 2.9) {
-            textBlock.setClassification("TextBlock Italic");
+            textBlock.setClassification(PageBlockType.PARAGRAPH_ITALIC);
         } else if (PositionUtils.isWithinBodyTextFrame(bodyTextFrame, textBlock)) {
-            textBlock.setClassification("TextBlock Unknown");
+            textBlock.setClassification(PageBlockType.PARAGRAPH_UNKNOWN);
         } else {
-            textBlock.setClassification("Other");
+            textBlock.setClassification(PageBlockType.OTHER);
         }
     }
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/service/PdfParsingService.java

@@ -9,16 +9,16 @@ import org.apache.pdfbox.pdmodel.PDPage;
 import org.apache.pdfbox.pdmodel.common.PDRectangle;
 import org.springframework.stereotype.Service;
 
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.ClassificationDocument;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.ClassificationPage;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.ClassificationTextBlock;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.TextPositionSequence;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.image.ClassifiedImage;
 import com.knecon.fforesight.service.layoutparser.processor.adapter.ImageServiceResponseAdapter;
+import com.knecon.fforesight.service.layoutparser.processor.adapter.model.table.TableCells;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.AbstractPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.ClassificationDocument;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.ClassificationPage;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.image.ClassifiedImage;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.table.CleanRulings;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.text.TextPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.text.TextPositionSequence;
 import com.knecon.fforesight.service.layoutparser.processor.classification.parsing.PDFLinesTextStripper;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.AbstractTextContainer;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.CleanRulings;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.CvParsedTableCell;
 
 import lombok.RequiredArgsConstructor;
 import lombok.SneakyThrows;
@@ -35,7 +35,7 @@ public class PdfParsingService {
     private final ImageServiceResponseAdapter imageServiceResponseAdapter;
 
 
-    public ClassificationDocument parseDocument(PDDocument originDocument, Map<Integer, List<CvParsedTableCell>> pdfTableCells, Map<Integer, List<ClassifiedImage>> pdfImages) {
+    public ClassificationDocument parseDocument(PDDocument originDocument, Map<Integer, List<TableCells>> pdfTableCells, Map<Integer, List<ClassifiedImage>> pdfImages) {
 
         ClassificationDocument document = new ClassificationDocument();
         List<ClassificationPage> classificationPages = new ArrayList<>();
@@ -56,7 +56,7 @@ public class PdfParsingService {
     @SneakyThrows
     private void parsePage(Map<Integer, List<ClassifiedImage>> pdfImages,
                            PDDocument pdDocument,
-                           Map<Integer, List<CvParsedTableCell>> pdfTableCells,
+                           Map<Integer, List<TableCells>> pdfTableCells,
                            ClassificationDocument document,
                            List<ClassificationPage> classificationPages,
                            int pageNumber) {
@@ -93,7 +93,7 @@ public class PdfParsingService {
             imageServiceResponseAdapter.findOcr(classificationPage);
         }
 
-        tableExtractionService.removeRedundantTableCells(cleanRulings, classificationPage);
+        tableExtractionService.extractTables(cleanRulings, classificationPage);
         buildPageStatistics(classificationPage);
         increaseDocumentStatistics(classificationPage, document);
 
@@ -115,12 +115,12 @@ public class PdfParsingService {
     private void buildPageStatistics(ClassificationPage classificationPage) {
 
         // Collect all statistics for the classificationPage, except from blocks inside tables, as tables will always be added to BodyTextFrame.
-        for (AbstractTextContainer textBlock : classificationPage.getTextBlocks()) {
-            if (textBlock instanceof ClassificationTextBlock) {
-                if (((ClassificationTextBlock) textBlock).getSequences() == null) {
+        for (AbstractPageBlock textBlock : classificationPage.getTextBlocks()) {
+            if (textBlock instanceof TextPageBlock) {
+                if (((TextPageBlock) textBlock).getSequences() == null) {
                     continue;
                 }
-                for (TextPositionSequence word : ((ClassificationTextBlock) textBlock).getSequences()) {
+                for (TextPositionSequence word : ((TextPageBlock) textBlock).getSequences()) {
                     classificationPage.getTextHeightCounter().add(word.getTextHeight());
                     classificationPage.getFontCounter().add(word.getFont());
                     classificationPage.getFontSizeCounter().add(word.getFontSize());
@@ -132,3 +132,5 @@ public class PdfParsingService {
     }
 
 }
+
+

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/service/RulingCleaningService.java

@@ -12,9 +12,9 @@ import java.util.Map;
 
 import org.springframework.stereotype.Service;
 
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.CleanRulings;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.CvParsedTableCell;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.Ruling;
+import com.knecon.fforesight.service.layoutparser.processor.adapter.model.table.TableCells;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.table.CleanRulings;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.table.Ruling;
 import com.knecon.fforesight.service.layoutparser.processor.classification.utils.DoubleComparisons;
 
 import lombok.RequiredArgsConstructor;
@@ -25,7 +25,7 @@ import lombok.extern.slf4j.Slf4j;
 @RequiredArgsConstructor
 public class RulingCleaningService {
 
-    public CleanRulings getCleanRulings(List<CvParsedTableCell> cvParsedTableCells, List<Ruling> rulings, float minCharWidth, float maxCharHeight) {
+    public CleanRulings getCleanRulings(List<TableCells> tableCells, List<Ruling> rulings, float minCharWidth, float maxCharHeight) {
 
         if (!rulings.isEmpty()) {
             snapPoints(rulings, minCharWidth, maxCharHeight);
@@ -38,7 +38,7 @@ public class RulingCleaningService {
             }
         }
         if (vrs.isEmpty()) {
-            vrs.addAll(extractVerticalRulings(cvParsedTableCells));
+            vrs.addAll(extractVerticalRulings(tableCells));
         }
         List<Ruling> verticalRulingLines = collapseOrientedRulings(vrs);
 
@@ -49,7 +49,7 @@ public class RulingCleaningService {
             }
         }
         if (hrs.isEmpty()) {
-            hrs.addAll(extractHorizontalRulings(cvParsedTableCells));
+            hrs.addAll(extractHorizontalRulings(tableCells));
         }
         List<Ruling> horizontalRulingLines = collapseOrientedRulings(hrs);
 
@@ -132,12 +132,12 @@ public class RulingCleaningService {
     }
 
 
-    private Collection<? extends Ruling> extractVerticalRulings(List<CvParsedTableCell> cvParsedTableCells) {
+    private Collection<? extends Ruling> extractVerticalRulings(List<TableCells> cvParsedTableCells) {
 
         List<Ruling> vrs = new ArrayList<>();
 
         if (cvParsedTableCells != null) {
-            for (CvParsedTableCell cvParsedTableCell : cvParsedTableCells) {
+            for (TableCells cvParsedTableCell : cvParsedTableCells) {
                 Ruling leftLine = createRuling(cvParsedTableCell.getX0(), cvParsedTableCell.getX0(), cvParsedTableCell.getY0(), cvParsedTableCell.getY1());
                 Ruling rightLine = createRuling(cvParsedTableCell.getX1(), cvParsedTableCell.getX1(), cvParsedTableCell.getY0(), cvParsedTableCell.getY1());
                 vrs.add(leftLine);
@@ -148,12 +148,12 @@ public class RulingCleaningService {
     }
 
 
-    private Collection<? extends Ruling> extractHorizontalRulings(List<CvParsedTableCell> cvParsedTableCells) {
+    private Collection<? extends Ruling> extractHorizontalRulings(List<TableCells> cvParsedTableCells) {
 
         List<Ruling> hrs = new ArrayList<>();
 
         if (cvParsedTableCells != null) {
-            for (CvParsedTableCell cvParsedTableCell : cvParsedTableCells) {
+            for (TableCells cvParsedTableCell : cvParsedTableCells) {
                 Ruling topLine = createRuling(cvParsedTableCell.getX0(), cvParsedTableCell.getX1(), cvParsedTableCell.getY1(), cvParsedTableCell.getY1());
                 Ruling baseLine = createRuling(cvParsedTableCell.getX0(), cvParsedTableCell.getX1(), cvParsedTableCell.getY0(), cvParsedTableCell.getY0());
                 hrs.add(topLine);

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/service/SectionsBuilderService.java

@@ -9,17 +9,18 @@ import java.util.stream.Collectors;
 
 import org.springframework.stereotype.Service;
 
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.ClassificationDocument;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.ClassificationFooter;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.ClassificationPage;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.ClassificationSection;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.ClassificationHeader;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.ClassificationTextBlock;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.UnclassifiedText;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.image.ClassifiedImage;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.AbstractTextContainer;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.TableCell;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.Table;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.AbstractPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.ClassificationDocument;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.ClassificationFooter;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.ClassificationHeader;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.ClassificationPage;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.ClassificationSection;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.PageBlockType;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.image.ClassifiedImage;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.table.Cell;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.table.TablePageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.text.TextPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.text.UnclassifiedText;
 
 import lombok.extern.slf4j.Slf4j;
 
@@ -29,44 +30,44 @@ public class SectionsBuilderService {
 
     public void buildSections(ClassificationDocument document) {
 
-        List<AbstractTextContainer> chunkWords = 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 (ClassificationPage classificationPage : document.getPages()) {
-            List<ClassificationTextBlock> header = new ArrayList<>();
-            List<ClassificationTextBlock> footer = new ArrayList<>();
-            List<ClassificationTextBlock> unclassifiedText = new ArrayList<>();
-            for (AbstractTextContainer current : classificationPage.getTextBlocks()) {
+        TablePageBlock previousTable = null;
+        for (ClassificationPage page : document.getPages()) {
+            List<TextPageBlock> header = new ArrayList<>();
+            List<TextPageBlock> footer = new ArrayList<>();
+            List<TextPageBlock> unclassifiedText = new ArrayList<>();
+            for (AbstractPageBlock current : page.getTextBlocks()) {
 
                 if (current.getClassification() == null) {
                     continue;
                 }
 
-                current.setPage(classificationPage.getPageNumber());
+                current.setPage(page.getPageNumber());
 
-                if (current.getClassification().equals("Header")) {
-                    header.add((ClassificationTextBlock) current);
+                if (current.getClassification().equals(PageBlockType.HEADER)) {
+                    header.add((TextPageBlock) current);
                     continue;
                 }
 
-                if (current.getClassification().equals("Footer")) {
-                    footer.add((ClassificationTextBlock) current);
+                if (current.getClassification().equals(PageBlockType.FOOTER)) {
+                    footer.add((TextPageBlock) current);
                     continue;
                 }
 
-                if (current.getClassification().equals("Other")) {
-                    unclassifiedText.add((ClassificationTextBlock) current);
+                if (current.getClassification().equals(PageBlockType.OTHER)) {
+                    unclassifiedText.add((TextPageBlock) current);
                     continue;
                 }
 
-                if (prev != null && current.getClassification().startsWith("H ") && !prev.getClassification().startsWith("H ") || !document.isHeadlines()) {
+                if (prev != null && current.getClassification().isHeadline() && !prev.getClassification().isHeadline() || !document.isHeadlines()) {
                     ClassificationSection chunkBlock = buildTextBlock(chunkWords, lastHeadline);
                     chunkBlock.setHeadline(lastHeadline);
                     if (document.isHeadlines()) {
@@ -78,7 +79,7 @@ public class SectionsBuilderService {
                         previousTable = chunkBlock.getTables().get(chunkBlock.getTables().size() - 1);
                     }
                 }
-                if (current instanceof Table table) {
+                if (current instanceof TablePageBlock table) {
                     // Distribute header information for subsequent tables
                     mergeTableMetadata(table, previousTable);
                     previousTable = table;
@@ -106,15 +107,14 @@ public class SectionsBuilderService {
         document.setHeaders(headers);
         document.setFooters(footers);
         document.setUnclassifiedTexts(unclassifiedTexts);
-        addImagesToSections(document);
     }
 
 
-    private void addImagesToSections(ClassificationDocument document) {
+    public void addImagesToSections(ClassificationDocument document) {
 
         Map<Integer, List<ClassificationSection>> sectionMap = new HashMap<>();
         for (ClassificationSection section : document.getSections()) {
-            for (AbstractTextContainer container : section.getPageBlocks()) {
+            for (AbstractPageBlock container : section.getPageBlocks()) {
 
                 List<ClassificationSection> sectionsOnPage = sectionMap.computeIfAbsent(container.getPage(), c -> new ArrayList<>());
                 if (sectionsOnPage.contains(section)) {
@@ -138,11 +138,11 @@ public class SectionsBuilderService {
             sectionMap.computeIfAbsent(1, x -> new ArrayList<>()).add(section);
         }
 
-        for (ClassificationPage classificationPage : document.getPages()) {
-            for (ClassifiedImage image : classificationPage.getImages()) {
-                List<ClassificationSection> sectionsOnPage = sectionMap.get(classificationPage.getPageNumber());
+        for (ClassificationPage page : document.getPages()) {
+            for (ClassifiedImage image : page.getImages()) {
+                List<ClassificationSection> sectionsOnPage = sectionMap.get(page.getPageNumber());
                 if (sectionsOnPage == null) {
-                    int i = classificationPage.getPageNumber();
+                    int i = page.getPageNumber();
                     while (sectionsOnPage == null) {
                         sectionsOnPage = sectionMap.get(i);
                         i--;
@@ -154,8 +154,8 @@ public class SectionsBuilderService {
                     Float xMax = null;
                     Float yMax = null;
 
-                    for (AbstractTextContainer abs : section.getPageBlocks()) {
-                        if (abs.getPage() != classificationPage.getPageNumber()) {
+                    for (AbstractPageBlock abs : section.getPageBlocks()) {
+                        if (abs.getPage() != page.getPageNumber()) {
                             continue;
                         }
 
@@ -212,23 +212,23 @@ public class SectionsBuilderService {
     }
 
 
-    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)) {
-            List<TableCell> previousTableNonHeaderRow = getRowWithNonHeaderCells(previousTable);
-            List<TableCell> tableNonHeaderRow = getRowWithNonHeaderCells(currentTable);
+            List<Cell> previousTableNonHeaderRow = getRowWithNonHeaderCells(previousTable);
+            List<Cell> tableNonHeaderRow = getRowWithNonHeaderCells(currentTable);
             // Allow merging of tables if header row is separated from first logical non-header row
             if (previousTableNonHeaderRow.isEmpty() && previousTable.getRowCount() == 1 && previousTable.getRows().get(0).size() == tableNonHeaderRow.size()) {
                 previousTableNonHeaderRow = previousTable.getRows().get(0).stream().map(cell -> {
-                    TableCell fakeCell = new TableCell(cell.getPoints()[0], cell.getPoints()[2]);
+                    Cell fakeCell = new Cell(cell.getPoints()[0], cell.getPoints()[2]);
                     fakeCell.setHeaderCells(Collections.singletonList(cell));
                     return fakeCell;
                 }).collect(Collectors.toList());
             }
             if (previousTableNonHeaderRow.size() == tableNonHeaderRow.size()) {
                 for (int i = currentTable.getRowCount() - 1; i >= 0; i--) { // Non header rows are most likely at bottom of table
-                    List<TableCell> row = currentTable.getRows().get(i);
+                    List<Cell> row = currentTable.getRows().get(i);
                     if (row.size() == tableNonHeaderRow.size() && row.stream().allMatch(cell -> cell.getHeaderCells().isEmpty())) {
                         for (int j = 0; j < row.size(); j++) {
                             row.get(j).setHeaderCells(previousTableNonHeaderRow.get(j).getHeaderCells());
@@ -240,52 +240,52 @@ public class SectionsBuilderService {
     }
 
 
-    private ClassificationSection buildTextBlock(List<AbstractTextContainer> wordBlockList, String lastHeadline) {
+    private ClassificationSection buildTextBlock(List<AbstractPageBlock> wordBlockList, String lastHeadline) {
 
         ClassificationSection section = new ClassificationSection();
 
-        for (AbstractTextContainer container : wordBlockList) {
-            if (container instanceof Table table) {
+        for (AbstractPageBlock container : wordBlockList) {
+            if (container instanceof TablePageBlock table) {
 
                 if (lastHeadline == null || lastHeadline.isEmpty()) {
                     table.setHeadline("Text in table");
                 } else {
-                    table.setHeadline("Table in: " + lastHeadline);
+                    table.setHeadline("TablePageBlock in: " + lastHeadline);
                 }
 
                 section.getPageBlocks().add(table);
                 continue;
             }
 
-            ClassificationTextBlock wordBlock = (ClassificationTextBlock) container;
+            TextPageBlock wordBlock = (TextPageBlock) container;
             section.getPageBlocks().add(wordBlock);
         }
         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 -> !cell.getHeaderCells().isEmpty())).findAny().isEmpty();
 
     }
 
 
-    private List<TableCell> 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<TableCell> row = table.getRows().get(i);
+            List<Cell> row = table.getRows().get(i);
             if (row.size() == 1) {
                 continue;
             }
             boolean allNonHeader = true;
-            for (TableCell cell : row) {
+            for (Cell cell : row) {
                 if (cell.isHeaderCell()) {
                     allNonHeader = false;
                     break;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/service/TableExtractionService.java

@@ -9,20 +9,18 @@ import java.util.Iterator;
 import java.util.List;
 import java.util.Map;
 import java.util.Set;
-import java.util.stream.Collectors;
 
 import org.springframework.stereotype.Service;
 
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.ClassificationPage;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.ClassificationTextBlock;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.TableCell;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.CleanRulings;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.Ruling;
-import com.knecon.fforesight.service.layoutparser.processor.classification.utils.QuickSort;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.AbstractPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.ClassificationPage;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.table.Cell;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.table.CleanRulings;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.table.Rectangle;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.table.Ruling;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.table.TablePageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.text.TextPageBlock;
 import com.knecon.fforesight.service.layoutparser.processor.classification.utils.DoubleComparisons;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.AbstractTextContainer;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.Rectangle;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.Table;
 
 @Service
 public class TableExtractionService {
@@ -68,28 +66,28 @@ public class TableExtractionService {
 
 
     /**
-     * Finds tables on a classificationPage and moves textblocks into cells of the found tables.
-     * Note: This algorithm uses Pdf Coordinate System where {0,0} rotated with the classificationPage rotation.
+     * Finds tables on a page and moves textblocks into cells of the found tables.
+     * Note: This algorithm uses Pdf Coordinate System where {0,0} rotated with the page rotation.
      * 0 -> LowerLeft
      * 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 classificationPage ClassificationPage object that contains textblocks and statistics.
+     * @param page         Page object that contains textblocks and statistics.
      */
-    public void removeRedundantTableCells(CleanRulings cleanRulings, ClassificationPage classificationPage) {
+    public void extractTables(CleanRulings cleanRulings, ClassificationPage page) {
 
-        List<TableCell> cells = findCells(cleanRulings.getHorizontal(), cleanRulings.getVertical());
+        List<Cell> cells = findCells(cleanRulings.getHorizontal(), cleanRulings.getVertical());
 
-        List<ClassificationTextBlock> toBeRemoved = new ArrayList<>();
+        List<TextPageBlock> toBeRemoved = new ArrayList<>();
 
-        for (AbstractTextContainer abstractTextContainer : classificationPage.getTextBlocks()) {
-            ClassificationTextBlock textBlock = (ClassificationTextBlock) abstractTextContainer;
-            for (TableCell cell : cells) {
-                if (cell.intersects(textBlock.getPdfMinX(),
+        for (AbstractPageBlock abstractPageBlock : page.getTextBlocks()) {
+            TextPageBlock textBlock = (TextPageBlock) abstractPageBlock;
+            for (Cell cell : cells) {
+                if (cell.hasMinimumSize() && cell.intersects(textBlock.getPdfMinX(),
                         textBlock.getPdfMinY(),
                         textBlock.getPdfMaxX() - textBlock.getPdfMinX(),
                         textBlock.getPdfMaxY() - textBlock.getPdfMinY())) {
@@ -101,44 +99,52 @@ public class TableExtractionService {
         }
 
         cells = new ArrayList<>(new HashSet<>(cells));
-        QuickSort.sort(cells, Rectangle.ILL_DEFINED_ORDER);
+        DoubleComparisons.sort(cells, Rectangle.ILL_DEFINED_ORDER);
 
-        List<Rectangle> spreadsheetAreas = findSpreadsheetsFromCells(cells).stream().filter(r -> r.getWidth() > 0f && r.getHeight() > 0f).collect(Collectors.toList());
+        List<Rectangle> spreadsheetAreas = findSpreadsheetsFromCells(cells).stream().filter(r -> r.getWidth() > 0f && r.getHeight() > 0f).toList();
 
-        List<Table> tables = new ArrayList<>();
+        List<TablePageBlock> tables = new ArrayList<>();
         for (Rectangle area : spreadsheetAreas) {
 
-            List<TableCell> overlappingCells = new ArrayList<>();
-            for (TableCell c : cells) {
-                if (c.intersects(area)) {
+            List<Cell> overlappingCells = new ArrayList<>();
+            for (Cell c : cells) {
+                if (c.hasMinimumSize() && c.intersects(area)) {
                     overlappingCells.add(c);
                 }
             }
-            tables.add(new Table(overlappingCells, area, classificationPage.getRotation()));
+            tables.add(new TablePageBlock(overlappingCells, area, page.getRotation()));
         }
 
-        for (Table table : tables) {
+        for (TablePageBlock table : tables) {
             int position = -1;
 
-            Iterator<AbstractTextContainer> itty = classificationPage.getTextBlocks().iterator();
+            Iterator<AbstractPageBlock> itty = page.getTextBlocks().iterator();
             while (itty.hasNext()) {
-                AbstractTextContainer textBlock = itty.next();
-                if (textBlock instanceof ClassificationTextBlock ? table.containsBlock((ClassificationTextBlock) textBlock) : table.contains(textBlock) && position == -1) {
-                    position = classificationPage.getTextBlocks().indexOf(textBlock);
+                AbstractPageBlock textBlock = itty.next();
+                if (textBlock instanceof TextPageBlock ? table.containsBlock((TextPageBlock) textBlock) : table.contains(textBlock) && position == -1) {
+                    position = page.getTextBlocks().indexOf(textBlock);
                 }
             }
             if (position != -1) {
-                classificationPage.getTextBlocks().add(position, table);
+                page.getTextBlocks().add(position, table);
             }
         }
 
-        classificationPage.getTextBlocks().removeAll(toBeRemoved);
+        page.getTextBlocks().removeAll(toBeRemoved);
     }
 
 
-    public List<TableCell> findCells(List<Ruling> horizontalRulingLines, List<Ruling> verticalRulingLines) {
+    public List<Cell> findCells(List<Ruling> horizontalRulingLines, List<Ruling> verticalRulingLines) {
 
-        List<TableCell> cellsFound = new ArrayList<>();
+        for (Ruling r : horizontalRulingLines) {
+            if (r.getX2() < r.getX1()) {
+                double a = r.getX2();
+                r.x2 = (float) r.getX1();
+                r.x1 = (float) a;
+            }
+        }
+
+        List<Cell> cellsFound = new ArrayList<>();
         Map<Point2D, Ruling[]> intersectionPoints = Ruling.findIntersections(horizontalRulingLines, verticalRulingLines);
         List<Point2D> intersectionPointsList = new ArrayList<>(intersectionPoints.keySet());
         intersectionPointsList.sort(POINT_COMPARATOR);
@@ -174,7 +180,7 @@ public class TableExtractionService {
                     Point2D btmRight = new Point2D.Float((float) yPoint.getX(), (float) xPoint.getY());
                     if (intersectionPoints.containsKey(btmRight) && intersectionPoints.get(btmRight)[0].equals(intersectionPoints.get(xPoint)[0]) && intersectionPoints.get(btmRight)[1].equals(
                             intersectionPoints.get(yPoint)[1])) {
-                        cellsFound.add(new TableCell(topLeft, btmRight));
+                        cellsFound.add(new Cell(topLeft, btmRight));
                         break outer;
                     }
                 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/utils/DoubleComparisons.java

@@ -1,6 +1,8 @@
 package com.knecon.fforesight.service.layoutparser.processor.classification.utils;
 
 import java.math.BigDecimal;
+import java.util.Comparator;
+import java.util.List;
 
 import lombok.experimental.UtilityClass;
 import lombok.extern.slf4j.Slf4j;
@@ -20,11 +22,22 @@ public final class DoubleComparisons {
 
 
     public static float round(double d, int decimalPlace) {
+
         BigDecimal bd = BigDecimal.valueOf(d);
         bd = bd.setScale(decimalPlace, BigDecimal.ROUND_HALF_UP);
         return bd.floatValue();
     }
 
 
+    public static <T> void sort(List<T> list, Comparator<? super T> comparator) {
+
+        try {
+            QuickSort.sort(list, comparator);
+        } catch (IllegalArgumentException e) {
+            // This should not happen since we use QuickSort from PDFBox
+            log.warn(e.getMessage());
+        }
+    }
+
 }
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/utils/PositionUtils.java

@@ -1,7 +1,7 @@
 package com.knecon.fforesight.service.layoutparser.processor.classification.utils;
 
 import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.Rectangle;
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.text.ClassificationTextBlock;
+import com.knecon.fforesight.service.layoutparser.processor.classification.model.text.TextPageBlock;
 
 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, ClassificationTextBlock textBlock) {
+    public boolean isWithinBodyTextFrame(Rectangle btf, TextPageBlock 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, ClassificationTextBlock textBlock, int rotation) {
+    public boolean isOverBodyTextFrame(Rectangle btf, TextPageBlock 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, ClassificationTextBlock textBlock, int rotation) {
+    public boolean isUnderBodyTextFrame(Rectangle btf, TextPageBlock 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, ClassificationTextBlock textBlock) {
+    public boolean isTouchingUnderBodyTextFrame(Rectangle btf, TextPageBlock textBlock) {
 
         //TODO Currently this is not working for rotated pages.
 
@@ -105,13 +107,13 @@ public final class PositionUtils {
     }
 
 
-    public float getHeightDifferenceBetweenChunkWordAndDocumentWord(ClassificationTextBlock textBlock, Float documentMostPopularWordHeight) {
+    public float getHeightDifferenceBetweenChunkWordAndDocumentWord(TextPageBlock textBlock, Float documentMostPopularWordHeight) {
 
         return textBlock.getMostPopularWordHeight() - documentMostPopularWordHeight;
     }
 
 
-    public Float getApproxLineCount(ClassificationTextBlock textBlock) {
+    public Float getApproxLineCount(TextPageBlock textBlock) {
 
         return textBlock.getHeight() / textBlock.getMostPopularWordHeight();
     }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/utils/RulingTextDirAdjustUtil.java

@@ -3,7 +3,7 @@ package com.knecon.fforesight.service.layoutparser.processor.classification.util
 import java.awt.geom.Line2D;
 import java.awt.geom.Point2D;
 
-import com.knecon.fforesight.service.layoutparser.processor.classification.dto.table.Ruling;
+import com.knecon.fforesight.service.layoutparser.processor.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) {

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/classification/utils/TextNormalizationUtilities.java

@@ -16,4 +16,16 @@ public final class TextNormalizationUtilities {
         return text.replaceAll("([^\\s\\d\\-]{2,500})[\\-\\u00AD]\\R", "$1");
     }
 
+
+    public static String removeLineBreaks(String text) {
+
+        return text.replaceAll("\n", " ");
+    }
+
+
+    public static String removeRepeatingWhitespaces(String text) {
+
+        return text.replaceAll(" {2}", " ");
+    }
+
 }