CLARI-002: fix some stuff with DocumentDataParser

* still todo, exlude semanticNodes inside TableCells

.gitlab-ci.yml

@@ -21,6 +21,5 @@ deploy:
       dotenv: version.env
   rules:
     - if: $CI_COMMIT_BRANCH == $CI_DEFAULT_BRANCH 
-    - if: $CI_COMMIT_BRANCH =~ /^feature/ && $CI_COMMIT_TAG == ""
     - if: $CI_COMMIT_BRANCH =~ /^release/
     - if: $CI_COMMIT_TAG

buildSrc/src/main/kotlin/com.knecon.fforesight.java-conventions.gradle.kts

@@ -42,19 +42,6 @@ tasks.jacocoTestReport {
 }
 
 allprojects {
-
-    tasks.withType<Javadoc> {
-        options {
-            this as StandardJavadocDocletOptions
-            addBooleanOption("Xdoclint:none", true)
-            addStringOption("Xmaxwarns", "1")
-        }
-    }
-
-    pmd {
-        setConsoleOutput(true)
-    }
-
     publishing {
         publications {
             create<MavenPublication>(name) {
@@ -77,7 +64,6 @@ java {
     withJavadocJar()
 }
 
-
 repositories {
     mavenLocal()
     mavenCentral()

layoutparser-service/layoutparser-service-internal-api/build.gradle.kts

@@ -7,5 +7,4 @@ description = "layoutparser-service-internal-api"
 
 dependencies {
     implementation("io.swagger.core.v3:swagger-annotations:2.2.15")
-    implementation("com.google.protobuf:protobuf-java-util:4.27.1")
 }

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/clarifynd/IndexData.java

@@ -0,0 +1,11 @@
+package com.knecon.fforesight.service.layoutparser.internal.api.data.clarifynd;
+
+import java.util.List;
+import java.util.Map;
+
+public class IndexData {
+
+    Map<String, String> identifier;
+    List<TextChunk> textChunks;
+
+}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/clarifynd/TextChunk.java

@@ -0,0 +1,18 @@
+package com.knecon.fforesight.service.layoutparser.internal.api.data.clarifynd;
+
+import java.util.List;
+
+import com.knecon.fforesight.service.layoutparser.internal.api.data.taas.Range;
+
+import lombok.AccessLevel;
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.experimental.FieldDefaults;
+
+@Builder
+@AllArgsConstructor
+@FieldDefaults(makeFinal = true, level = AccessLevel.PRIVATE)
+public class TextChunk {
+
+    String text;
+}

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

@@ -2,11 +2,6 @@ package com.knecon.fforesight.service.layoutparser.internal.api.data.redaction;
 
 import java.io.Serializable;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentPageProto.AllDocumentPages;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentPositionDataProto.AllDocumentPositionData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentStructureProto.DocumentStructure;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentTextDataProto.AllDocumentTextData;
-
 import io.swagger.v3.oas.annotations.media.Schema;
 import lombok.AccessLevel;
 import lombok.AllArgsConstructor;
@@ -18,26 +13,16 @@ import lombok.experimental.FieldDefaults;
 @Builder
 @AllArgsConstructor
 @FieldDefaults(makeFinal = true, level = AccessLevel.PRIVATE)
-@Schema(description = "Object containing the complete document layout parsing information. It is split into 4 categories, structure, text, positions and pages: "
-                      + "The document tree structure of SemanticNodes such as Section, Paragraph, Headline, etc. "
-                      + "The text, which is stored as separate blocks of data. "
-                      + "The text positions, which are also stored as separate blocks. The Blocks are equal to the text blocks in length and order. "
-                      + "The page information.")
+@Schema(description = "Object containing the complete document layout parsing information. It is split into 4 categories, structure, text, positions and pages: " + "The document tree structure of SemanticNodes such as Section, Paragraph, Headline, etc. " + "The text, which is stored as separate blocks of data. " + "The text positions, which are also stored as separate blocks. The Blocks are equal to the text blocks in length and order. " + "The page information.")
 public class DocumentData implements Serializable {
 
     @Schema(description = "Contains information about the document's pages.")
-    AllDocumentPages documentPages;
+    DocumentPage[] documentPages;
     @Schema(description = "Contains information about the document's text.")
-    AllDocumentTextData documentTextData;
+    DocumentTextData[] documentTextData;
     @Schema(description = "Contains information about the document's text positions.")
-    AllDocumentPositionData documentPositions;
+    DocumentPositionData[] documentPositions;
     @Schema(description = "Contains information about the document's semantic structure.")
-    DocumentStructureWrapper documentStructureWrapper;
-
-
-    public DocumentStructure getDocumentStructure() {
-
-        return documentStructureWrapper.getDocumentStructure();
-    }
+    DocumentStructure documentStructure;
 
 }

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

@@ -10,7 +10,6 @@ import lombok.Data;
 import lombok.NoArgsConstructor;
 import lombok.experimental.FieldDefaults;
 
-@Deprecated
 @Data
 @Builder
 @NoArgsConstructor

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

@@ -10,7 +10,6 @@ import lombok.Data;
 import lombok.NoArgsConstructor;
 import lombok.experimental.FieldDefaults;
 
-@Deprecated
 @Data
 @Builder
 @NoArgsConstructor

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

@@ -16,7 +16,6 @@ import lombok.Data;
 import lombok.NoArgsConstructor;
 import lombok.experimental.FieldDefaults;
 
-@Deprecated
 @Data
 @Builder
 @NoArgsConstructor
@@ -44,8 +43,6 @@ public class DocumentStructure implements Serializable {
         public static final String POSITION = "position";
         public static final String ID = "id";
 
-        public static final String REPRESENTATION_HASH = "representationHash";
-
     }
 
     @Schema(description = "Object containing the extra field names, a table cell has in its properties field.")
@@ -70,25 +67,11 @@ public class DocumentStructure implements Serializable {
 
     public static Rectangle2D parseRectangle2D(String bBox) {
 
-        List<Float> floats = Arrays.stream(bBox.split(RECTANGLE_DELIMITER))
-                .map(Float::parseFloat)
-                .toList();
+        List<Float> floats = Arrays.stream(bBox.split(RECTANGLE_DELIMITER)).map(Float::parseFloat).toList();
         return new Rectangle2D.Float(floats.get(0), floats.get(1), floats.get(2), floats.get(3));
     }
 
 
-    public static double[] parseRepresentationVector(String representationHash) {
-
-        String[] stringArray = representationHash.split("[,\\s]+");
-        double[] doubleArray = new double[stringArray.length];
-        for (int i = 0; i < stringArray.length; i++) {
-            doubleArray[i] = Double.parseDouble(stringArray[i]);
-        }
-
-        return doubleArray;
-    }
-
-
     public EntryData get(List<Integer> tocId) {
 
         if (tocId.isEmpty()) {
@@ -104,24 +87,19 @@ public class DocumentStructure implements Serializable {
 
     public Stream<EntryData> streamAllEntries() {
 
-        return Stream.concat(Stream.of(root), root.children.stream())
-                .flatMap(DocumentStructure::flatten);
+        return Stream.concat(Stream.of(root), root.children.stream()).flatMap(DocumentStructure::flatten);
     }
 
 
     public String toString() {
 
-        return String.join("\n",
-                           streamAllEntries().map(EntryData::toString)
-                                   .toList());
+        return String.join("\n", streamAllEntries().map(EntryData::toString).toList());
     }
 
 
     private static Stream<EntryData> flatten(EntryData entry) {
 
-        return Stream.concat(Stream.of(entry),
-                             entry.children.stream()
-                                     .flatMap(DocumentStructure::flatten));
+        return Stream.concat(Stream.of(entry), entry.children.stream().flatMap(DocumentStructure::flatten));
     }
 
 

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

@@ -1,799 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.data.redaction;
-
-// Generated by the protocol buffer compiler.  DO NOT EDIT!
-// NO CHECKED-IN PROTOBUF GENCODE
-// source: DocumentStructure.proto
-// Protobuf Java Version: 4.27.1
-@SuppressWarnings("all")
-public final class DocumentStructureProto {
-
-    private DocumentStructureProto() {}
-
-
-    static {
-        com.google.protobuf.RuntimeVersion.validateProtobufGencodeVersion(com.google.protobuf.RuntimeVersion.RuntimeDomain.PUBLIC,
-                /* major= */ 4,
-                /* minor= */ 27,
-                /* patch= */ 1,
-                /* suffix= */ "", DocumentStructureProto.class.getName());
-    }
-
-    public static void registerAllExtensions(com.google.protobuf.ExtensionRegistryLite registry) {
-
-    }
-
-
-    public static void registerAllExtensions(com.google.protobuf.ExtensionRegistry registry) {
-
-        registerAllExtensions((com.google.protobuf.ExtensionRegistryLite) registry);
-    }
-
-
-    public interface DocumentStructureOrBuilder extends
-            // @@protoc_insertion_point(interface_extends:DocumentStructure)
-            com.google.protobuf.MessageOrBuilder {
-
-        /**
-         * <pre>
-         * The root EntryData represents the Document.
-         * </pre>
-         *
-         * <code>.EntryData root = 1;</code>
-         *
-         * @return Whether the root field is set.
-         */
-        boolean hasRoot();
-
-
-        /**
-         * <pre>
-         * The root EntryData represents the Document.
-         * </pre>
-         *
-         * <code>.EntryData root = 1;</code>
-         *
-         * @return The root.
-         */
-        EntryDataProto.EntryData getRoot();
-
-
-        /**
-         * <pre>
-         * The root EntryData represents the Document.
-         * </pre>
-         *
-         * <code>.EntryData root = 1;</code>
-         */
-        EntryDataProto.EntryDataOrBuilder getRootOrBuilder();
-
-    }
-
-    /**
-     * Protobuf type {@code DocumentStructure}
-     */
-    public static final class DocumentStructure extends com.google.protobuf.GeneratedMessage implements
-            // @@protoc_insertion_point(message_implements:DocumentStructure)
-            DocumentStructureOrBuilder {
-
-        private static final long serialVersionUID = 0L;
-
-        static {
-            com.google.protobuf.RuntimeVersion.validateProtobufGencodeVersion(com.google.protobuf.RuntimeVersion.RuntimeDomain.PUBLIC,
-                    /* major= */ 4,
-                    /* minor= */ 27,
-                    /* patch= */ 1,
-                    /* suffix= */ "", DocumentStructure.class.getName());
-        }
-
-        // Use DocumentStructure.newBuilder() to construct.
-        private DocumentStructure(com.google.protobuf.GeneratedMessage.Builder<?> builder) {
-
-            super(builder);
-        }
-
-
-        private DocumentStructure() {
-
-        }
-
-
-        public static final com.google.protobuf.Descriptors.Descriptor getDescriptor() {
-
-            return DocumentStructureProto.internal_static_DocumentStructure_descriptor;
-        }
-
-
-        @Override
-        protected FieldAccessorTable internalGetFieldAccessorTable() {
-
-            return DocumentStructureProto.internal_static_DocumentStructure_fieldAccessorTable.ensureFieldAccessorsInitialized(DocumentStructure.class, Builder.class);
-        }
-
-
-        private int bitField0_;
-        public static final int ROOT_FIELD_NUMBER = 1;
-        private EntryDataProto.EntryData root_;
-
-
-        /**
-         * <pre>
-         * The root EntryData represents the Document.
-         * </pre>
-         *
-         * <code>.EntryData root = 1;</code>
-         *
-         * @return Whether the root field is set.
-         */
-        @Override
-        public boolean hasRoot() {
-
-            return ((bitField0_ & 0x00000001) != 0);
-        }
-
-
-        /**
-         * <pre>
-         * The root EntryData represents the Document.
-         * </pre>
-         *
-         * <code>.EntryData root = 1;</code>
-         *
-         * @return The root.
-         */
-        @Override
-        public EntryDataProto.EntryData getRoot() {
-
-            return root_ == null ? EntryDataProto.EntryData.getDefaultInstance() : root_;
-        }
-
-
-        /**
-         * <pre>
-         * The root EntryData represents the Document.
-         * </pre>
-         *
-         * <code>.EntryData root = 1;</code>
-         */
-        @Override
-        public EntryDataProto.EntryDataOrBuilder getRootOrBuilder() {
-
-            return root_ == null ? EntryDataProto.EntryData.getDefaultInstance() : root_;
-        }
-
-
-        private byte memoizedIsInitialized = -1;
-
-
-        @Override
-        public final boolean isInitialized() {
-
-            byte isInitialized = memoizedIsInitialized;
-          if (isInitialized == 1) {
-            return true;
-          }
-          if (isInitialized == 0) {
-            return false;
-          }
-
-            memoizedIsInitialized = 1;
-            return true;
-        }
-
-
-        @Override
-        public void writeTo(com.google.protobuf.CodedOutputStream output) throws java.io.IOException {
-
-            if (((bitField0_ & 0x00000001) != 0)) {
-                output.writeMessage(1, getRoot());
-            }
-            getUnknownFields().writeTo(output);
-        }
-
-
-        @Override
-        public int getSerializedSize() {
-
-            int size = memoizedSize;
-          if (size != -1) {
-            return size;
-          }
-
-            size = 0;
-            if (((bitField0_ & 0x00000001) != 0)) {
-                size += com.google.protobuf.CodedOutputStream.computeMessageSize(1, getRoot());
-            }
-            size += getUnknownFields().getSerializedSize();
-            memoizedSize = size;
-            return size;
-        }
-
-
-        @Override
-        public boolean equals(final Object obj) {
-
-            if (obj == this) {
-                return true;
-            }
-            if (!(obj instanceof DocumentStructure)) {
-                return super.equals(obj);
-            }
-            DocumentStructure other = (DocumentStructure) obj;
-
-          if (hasRoot() != other.hasRoot()) {
-            return false;
-          }
-            if (hasRoot()) {
-              if (!getRoot().equals(other.getRoot())) {
-                return false;
-              }
-            }
-          if (!getUnknownFields().equals(other.getUnknownFields())) {
-            return false;
-          }
-            return true;
-        }
-
-
-        @Override
-        public int hashCode() {
-
-            if (memoizedHashCode != 0) {
-                return memoizedHashCode;
-            }
-            int hash = 41;
-            hash = (19 * hash) + getDescriptor().hashCode();
-            if (hasRoot()) {
-                hash = (37 * hash) + ROOT_FIELD_NUMBER;
-                hash = (53 * hash) + getRoot().hashCode();
-            }
-            hash = (29 * hash) + getUnknownFields().hashCode();
-            memoizedHashCode = hash;
-            return hash;
-        }
-
-
-        public static DocumentStructure parseFrom(java.nio.ByteBuffer data) throws com.google.protobuf.InvalidProtocolBufferException {
-
-            return PARSER.parseFrom(data);
-        }
-
-
-        public static DocumentStructure parseFrom(java.nio.ByteBuffer data,
-                                                  com.google.protobuf.ExtensionRegistryLite extensionRegistry) throws com.google.protobuf.InvalidProtocolBufferException {
-
-            return PARSER.parseFrom(data, extensionRegistry);
-        }
-
-
-        public static DocumentStructure parseFrom(com.google.protobuf.ByteString data) throws com.google.protobuf.InvalidProtocolBufferException {
-
-            return PARSER.parseFrom(data);
-        }
-
-
-        public static DocumentStructure parseFrom(com.google.protobuf.ByteString data,
-                                                  com.google.protobuf.ExtensionRegistryLite extensionRegistry) throws com.google.protobuf.InvalidProtocolBufferException {
-
-            return PARSER.parseFrom(data, extensionRegistry);
-        }
-
-
-        public static DocumentStructure parseFrom(byte[] data) throws com.google.protobuf.InvalidProtocolBufferException {
-
-            return PARSER.parseFrom(data);
-        }
-
-
-        public static DocumentStructure parseFrom(byte[] data,
-                                                  com.google.protobuf.ExtensionRegistryLite extensionRegistry) throws com.google.protobuf.InvalidProtocolBufferException {
-
-            return PARSER.parseFrom(data, extensionRegistry);
-        }
-
-
-        public static DocumentStructure parseFrom(java.io.InputStream input) throws java.io.IOException {
-
-            return com.google.protobuf.GeneratedMessage.parseWithIOException(PARSER, input);
-        }
-
-
-        public static DocumentStructure parseFrom(java.io.InputStream input, com.google.protobuf.ExtensionRegistryLite extensionRegistry) throws java.io.IOException {
-
-            return com.google.protobuf.GeneratedMessage.parseWithIOException(PARSER, input, extensionRegistry);
-        }
-
-
-        public static DocumentStructure parseDelimitedFrom(java.io.InputStream input) throws java.io.IOException {
-
-            return com.google.protobuf.GeneratedMessage.parseDelimitedWithIOException(PARSER, input);
-        }
-
-
-        public static DocumentStructure parseDelimitedFrom(java.io.InputStream input, com.google.protobuf.ExtensionRegistryLite extensionRegistry) throws java.io.IOException {
-
-            return com.google.protobuf.GeneratedMessage.parseDelimitedWithIOException(PARSER, input, extensionRegistry);
-        }
-
-
-        public static DocumentStructure parseFrom(com.google.protobuf.CodedInputStream input) throws java.io.IOException {
-
-            return com.google.protobuf.GeneratedMessage.parseWithIOException(PARSER, input);
-        }
-
-
-        public static DocumentStructure parseFrom(com.google.protobuf.CodedInputStream input,
-                                                  com.google.protobuf.ExtensionRegistryLite extensionRegistry) throws java.io.IOException {
-
-            return com.google.protobuf.GeneratedMessage.parseWithIOException(PARSER, input, extensionRegistry);
-        }
-
-
-        @Override
-        public Builder newBuilderForType() {return newBuilder();}
-
-
-        public static Builder newBuilder() {
-
-            return DEFAULT_INSTANCE.toBuilder();
-        }
-
-
-        public static Builder newBuilder(DocumentStructure prototype) {
-
-            return DEFAULT_INSTANCE.toBuilder().mergeFrom(prototype);
-        }
-
-
-        @Override
-        public Builder toBuilder() {
-
-            return this == DEFAULT_INSTANCE ? new Builder() : new Builder().mergeFrom(this);
-        }
-
-
-        @Override
-        protected Builder newBuilderForType(BuilderParent parent) {
-
-            Builder builder = new Builder(parent);
-            return builder;
-        }
-
-
-        /**
-         * Protobuf type {@code DocumentStructure}
-         */
-        public static final class Builder extends com.google.protobuf.GeneratedMessage.Builder<Builder> implements
-                // @@protoc_insertion_point(builder_implements:DocumentStructure)
-                DocumentStructureOrBuilder {
-
-            public static final com.google.protobuf.Descriptors.Descriptor getDescriptor() {
-
-                return DocumentStructureProto.internal_static_DocumentStructure_descriptor;
-            }
-
-
-            @Override
-            protected FieldAccessorTable internalGetFieldAccessorTable() {
-
-                return DocumentStructureProto.internal_static_DocumentStructure_fieldAccessorTable.ensureFieldAccessorsInitialized(DocumentStructure.class, Builder.class);
-            }
-
-
-            // Construct using DocumentStructureOuterClass.DocumentStructure.newBuilder()
-            private Builder() {
-
-                maybeForceBuilderInitialization();
-            }
-
-
-            private Builder(BuilderParent parent) {
-
-                super(parent);
-                maybeForceBuilderInitialization();
-            }
-
-
-            private void maybeForceBuilderInitialization() {
-
-                if (com.google.protobuf.GeneratedMessage.alwaysUseFieldBuilders) {
-                    getRootFieldBuilder();
-                }
-            }
-
-
-            @Override
-            public Builder clear() {
-
-                super.clear();
-                bitField0_ = 0;
-                root_ = null;
-                if (rootBuilder_ != null) {
-                    rootBuilder_.dispose();
-                    rootBuilder_ = null;
-                }
-                return this;
-            }
-
-
-            @Override
-            public com.google.protobuf.Descriptors.Descriptor getDescriptorForType() {
-
-                return DocumentStructureProto.internal_static_DocumentStructure_descriptor;
-            }
-
-
-            @Override
-            public DocumentStructure getDefaultInstanceForType() {
-
-                return DocumentStructure.getDefaultInstance();
-            }
-
-
-            @Override
-            public DocumentStructure build() {
-
-                DocumentStructure result = buildPartial();
-                if (!result.isInitialized()) {
-                    throw newUninitializedMessageException(result);
-                }
-                return result;
-            }
-
-
-            @Override
-            public DocumentStructure buildPartial() {
-
-                DocumentStructure result = new DocumentStructure(this);
-                if (bitField0_ != 0) {
-                    buildPartial0(result);
-                }
-                onBuilt();
-                return result;
-            }
-
-
-            private void buildPartial0(DocumentStructure result) {
-
-                int from_bitField0_ = bitField0_;
-                int to_bitField0_ = 0;
-                if (((from_bitField0_ & 0x00000001) != 0)) {
-                    result.root_ = rootBuilder_ == null ? root_ : rootBuilder_.build();
-                    to_bitField0_ |= 0x00000001;
-                }
-                result.bitField0_ |= to_bitField0_;
-            }
-
-
-            @Override
-            public Builder mergeFrom(com.google.protobuf.Message other) {
-
-                if (other instanceof DocumentStructure) {
-                    return mergeFrom((DocumentStructure) other);
-                } else {
-                    super.mergeFrom(other);
-                    return this;
-                }
-            }
-
-
-            public Builder mergeFrom(DocumentStructure other) {
-
-              if (other == DocumentStructure.getDefaultInstance()) {
-                return this;
-              }
-                if (other.hasRoot()) {
-                    mergeRoot(other.getRoot());
-                }
-                this.mergeUnknownFields(other.getUnknownFields());
-                onChanged();
-                return this;
-            }
-
-
-            @Override
-            public final boolean isInitialized() {
-
-                return true;
-            }
-
-
-            @Override
-            public Builder mergeFrom(com.google.protobuf.CodedInputStream input, com.google.protobuf.ExtensionRegistryLite extensionRegistry) throws java.io.IOException {
-
-                if (extensionRegistry == null) {
-                    throw new NullPointerException();
-                }
-                try {
-                    boolean done = false;
-                    while (!done) {
-                        int tag = input.readTag();
-                        switch (tag) {
-                            case 0:
-                                done = true;
-                                break;
-                            case 10: {
-                                input.readMessage(getRootFieldBuilder().getBuilder(), extensionRegistry);
-                                bitField0_ |= 0x00000001;
-                                break;
-                            } // case 10
-                            default: {
-                                if (!super.parseUnknownField(input, extensionRegistry, tag)) {
-                                    done = true; // was an endgroup tag
-                                }
-                                break;
-                            } // default:
-                        } // switch (tag)
-                    } // while (!done)
-                } catch (com.google.protobuf.InvalidProtocolBufferException e) {
-                    throw e.unwrapIOException();
-                } finally {
-                    onChanged();
-                } // finally
-                return this;
-            }
-
-
-            private int bitField0_;
-
-            private EntryDataProto.EntryData root_;
-            private com.google.protobuf.SingleFieldBuilder<EntryDataProto.EntryData, EntryDataProto.EntryData.Builder, EntryDataProto.EntryDataOrBuilder> rootBuilder_;
-
-
-            /**
-             * <pre>
-             * The root EntryData represents the Document.
-             * </pre>
-             *
-             * <code>.EntryData root = 1;</code>
-             *
-             * @return Whether the root field is set.
-             */
-            public boolean hasRoot() {
-
-                return ((bitField0_ & 0x00000001) != 0);
-            }
-
-
-            /**
-             * <pre>
-             * The root EntryData represents the Document.
-             * </pre>
-             *
-             * <code>.EntryData root = 1;</code>
-             *
-             * @return The root.
-             */
-            public EntryDataProto.EntryData getRoot() {
-
-                if (rootBuilder_ == null) {
-                    return root_ == null ? EntryDataProto.EntryData.getDefaultInstance() : root_;
-                } else {
-                    return rootBuilder_.getMessage();
-                }
-            }
-
-
-            /**
-             * <pre>
-             * The root EntryData represents the Document.
-             * </pre>
-             *
-             * <code>.EntryData root = 1;</code>
-             */
-            public Builder setRoot(EntryDataProto.EntryData value) {
-
-                if (rootBuilder_ == null) {
-                    if (value == null) {
-                        throw new NullPointerException();
-                    }
-                    root_ = value;
-                } else {
-                    rootBuilder_.setMessage(value);
-                }
-                bitField0_ |= 0x00000001;
-                onChanged();
-                return this;
-            }
-
-
-            /**
-             * <pre>
-             * The root EntryData represents the Document.
-             * </pre>
-             *
-             * <code>.EntryData root = 1;</code>
-             */
-            public Builder setRoot(EntryDataProto.EntryData.Builder builderForValue) {
-
-                if (rootBuilder_ == null) {
-                    root_ = builderForValue.build();
-                } else {
-                    rootBuilder_.setMessage(builderForValue.build());
-                }
-                bitField0_ |= 0x00000001;
-                onChanged();
-                return this;
-            }
-
-
-            /**
-             * <pre>
-             * The root EntryData represents the Document.
-             * </pre>
-             *
-             * <code>.EntryData root = 1;</code>
-             */
-            public Builder mergeRoot(EntryDataProto.EntryData value) {
-
-                if (rootBuilder_ == null) {
-                    if (((bitField0_ & 0x00000001) != 0) && root_ != null && root_ != EntryDataProto.EntryData.getDefaultInstance()) {
-                        getRootBuilder().mergeFrom(value);
-                    } else {
-                        root_ = value;
-                    }
-                } else {
-                    rootBuilder_.mergeFrom(value);
-                }
-                if (root_ != null) {
-                    bitField0_ |= 0x00000001;
-                    onChanged();
-                }
-                return this;
-            }
-
-
-            /**
-             * <pre>
-             * The root EntryData represents the Document.
-             * </pre>
-             *
-             * <code>.EntryData root = 1;</code>
-             */
-            public Builder clearRoot() {
-
-                bitField0_ = (bitField0_ & ~0x00000001);
-                root_ = null;
-                if (rootBuilder_ != null) {
-                    rootBuilder_.dispose();
-                    rootBuilder_ = null;
-                }
-                onChanged();
-                return this;
-            }
-
-
-            /**
-             * <pre>
-             * The root EntryData represents the Document.
-             * </pre>
-             *
-             * <code>.EntryData root = 1;</code>
-             */
-            public EntryDataProto.EntryData.Builder getRootBuilder() {
-
-                bitField0_ |= 0x00000001;
-                onChanged();
-                return getRootFieldBuilder().getBuilder();
-            }
-
-
-            /**
-             * <pre>
-             * The root EntryData represents the Document.
-             * </pre>
-             *
-             * <code>.EntryData root = 1;</code>
-             */
-            public EntryDataProto.EntryDataOrBuilder getRootOrBuilder() {
-
-                if (rootBuilder_ != null) {
-                    return rootBuilder_.getMessageOrBuilder();
-                } else {
-                    return root_ == null ? EntryDataProto.EntryData.getDefaultInstance() : root_;
-                }
-            }
-
-
-            /**
-             * <pre>
-             * The root EntryData represents the Document.
-             * </pre>
-             *
-             * <code>.EntryData root = 1;</code>
-             */
-            private com.google.protobuf.SingleFieldBuilder<EntryDataProto.EntryData, EntryDataProto.EntryData.Builder, EntryDataProto.EntryDataOrBuilder> getRootFieldBuilder() {
-
-                if (rootBuilder_ == null) {
-                    rootBuilder_ = new com.google.protobuf.SingleFieldBuilder<EntryDataProto.EntryData, EntryDataProto.EntryData.Builder, EntryDataProto.EntryDataOrBuilder>(getRoot(),
-                                                                                                                                                                             getParentForChildren(),
-                                                                                                                                                                             isClean());
-                    root_ = null;
-                }
-                return rootBuilder_;
-            }
-
-            // @@protoc_insertion_point(builder_scope:DocumentStructure)
-        }
-
-        // @@protoc_insertion_point(class_scope:DocumentStructure)
-        private static final DocumentStructure DEFAULT_INSTANCE;
-
-        static {
-            DEFAULT_INSTANCE = new DocumentStructure();
-        }
-
-        public static DocumentStructure getDefaultInstance() {
-
-            return DEFAULT_INSTANCE;
-        }
-
-
-        private static final com.google.protobuf.Parser<DocumentStructure> PARSER = new com.google.protobuf.AbstractParser<DocumentStructure>() {
-            @Override
-            public DocumentStructure parsePartialFrom(com.google.protobuf.CodedInputStream input,
-                                                      com.google.protobuf.ExtensionRegistryLite extensionRegistry) throws com.google.protobuf.InvalidProtocolBufferException {
-
-                Builder builder = newBuilder();
-                try {
-                    builder.mergeFrom(input, extensionRegistry);
-                } catch (com.google.protobuf.InvalidProtocolBufferException e) {
-                    throw e.setUnfinishedMessage(builder.buildPartial());
-                } catch (com.google.protobuf.UninitializedMessageException e) {
-                    throw e.asInvalidProtocolBufferException().setUnfinishedMessage(builder.buildPartial());
-                } catch (java.io.IOException e) {
-                    throw new com.google.protobuf.InvalidProtocolBufferException(e).setUnfinishedMessage(builder.buildPartial());
-                }
-                return builder.buildPartial();
-            }
-        };
-
-
-        public static com.google.protobuf.Parser<DocumentStructure> parser() {
-
-            return PARSER;
-        }
-
-
-        @Override
-        public com.google.protobuf.Parser<DocumentStructure> getParserForType() {
-
-            return PARSER;
-        }
-
-
-        @Override
-        public DocumentStructure getDefaultInstanceForType() {
-
-            return DEFAULT_INSTANCE;
-        }
-
-    }
-
-    private static final com.google.protobuf.Descriptors.Descriptor internal_static_DocumentStructure_descriptor;
-    private static final com.google.protobuf.GeneratedMessage.FieldAccessorTable internal_static_DocumentStructure_fieldAccessorTable;
-
-
-    public static com.google.protobuf.Descriptors.FileDescriptor getDescriptor() {
-
-        return descriptor;
-    }
-
-
-    private static com.google.protobuf.Descriptors.FileDescriptor descriptor;
-
-    static {
-        String[] descriptorData = {"\n\027DocumentStructure.proto\032\017EntryData.pro"
-                                   + "to\"-\n\021DocumentStructure\022\030\n\004root\030\001 \001(\0132\n."
-                                   + "EntryDatab\006proto3"};
-        descriptor = com.google.protobuf.Descriptors.FileDescriptor.internalBuildGeneratedFileFrom(descriptorData,
-                                                                                                   new com.google.protobuf.Descriptors.FileDescriptor[]{EntryDataProto.getDescriptor(),});
-        internal_static_DocumentStructure_descriptor = getDescriptor().getMessageTypes()
-                .get(0);
-        internal_static_DocumentStructure_fieldAccessorTable = new com.google.protobuf.GeneratedMessage.FieldAccessorTable(internal_static_DocumentStructure_descriptor,
-                                                                                                                           new String[]{"Root",});
-        descriptor.resolveAllFeaturesImmutable();
-        EntryDataProto.getDescriptor();
-    }
-
-    // @@protoc_insertion_point(outer_class_scope)
-}

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

@@ -1,126 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.data.redaction;
-
-import static com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentStructureProto.DocumentStructure;
-
-import java.awt.geom.Rectangle2D;
-import java.io.ObjectStreamException;
-import java.io.Serializable;
-import java.util.Arrays;
-import java.util.List;
-import java.util.stream.Stream;
-
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.EntryDataProto.EntryData;
-
-import io.swagger.v3.oas.annotations.media.Schema;
-import lombok.AllArgsConstructor;
-import lombok.Getter;
-
-@Getter
-@AllArgsConstructor
-public class DocumentStructureWrapper implements Serializable {
-
-    private final DocumentStructure documentStructure;
-
-
-    @Schema(description = "Object containing the extra field names, a table has in its properties field.")
-    public static class TableProperties implements Serializable {
-
-        public static final String NUMBER_OF_ROWS = "numberOfRows";
-        public static final String NUMBER_OF_COLS = "numberOfCols";
-
-    }
-
-    @Schema(description = "Object containing the extra field names, an Image has in its properties field.")
-    public static class ImageProperties implements Serializable {
-
-        public static final String TRANSPARENT = "transparent";
-        public static final String IMAGE_TYPE = "imageType";
-        public static final String POSITION = "position";
-        public static final String ID = "id";
-
-        public static final String REPRESENTATION_HASH = "representationHash";
-
-    }
-
-    @Schema(description = "Object containing the extra field names, a table cell has in its properties field.")
-    public static class TableCellProperties implements Serializable {
-
-        public static final String B_BOX = "bBox";
-        public static final String ROW = "row";
-        public static final String COL = "col";
-        public static final String HEADER = "header";
-
-    }
-
-    @Schema(description = "Object containing the extra field names, a duplicate paragraph has in its properties field.")
-    public static class DuplicateParagraphProperties implements Serializable {
-
-        public static final String UNSORTED_TEXTBLOCK_ID = "utbid";
-
-    }
-
-    public static final String RECTANGLE_DELIMITER = ";";
-
-
-    public static Rectangle2D parseRectangle2D(String bBox) {
-
-        List<Float> floats = Arrays.stream(bBox.split(RECTANGLE_DELIMITER))
-                .map(Float::parseFloat)
-                .toList();
-        return new Rectangle2D.Float(floats.get(0), floats.get(1), floats.get(2), floats.get(3));
-    }
-
-
-    public static double[] parseRepresentationVector(String representationHash) {
-
-        String[] stringArray = representationHash.split("[,\\s]+");
-        double[] doubleArray = new double[stringArray.length];
-        for (int i = 0; i < stringArray.length; i++) {
-            doubleArray[i] = Double.parseDouble(stringArray[i]);
-        }
-
-        return doubleArray;
-    }
-
-
-    public EntryData get(List<Integer> tocId) {
-
-        if (tocId.isEmpty()) {
-            return documentStructure.getRoot();
-        }
-        EntryData entry = documentStructure.getRoot().getChildrenList()
-                .get(tocId.get(0));
-        for (int id : tocId.subList(1, tocId.size())) {
-            entry = entry.getChildrenList()
-                    .get(id);
-        }
-        return entry;
-    }
-
-
-    public Stream<EntryData> streamAllEntries() {
-
-        return Stream.concat(Stream.of(documentStructure.getRoot()),
-                             documentStructure.getRoot().getChildrenList()
-                                     .stream())
-                .flatMap(DocumentStructureWrapper::flatten);
-    }
-
-
-    public String toString() {
-
-        return String.join("\n",
-                           streamAllEntries().map(EntryData::toString)
-                                   .toList());
-    }
-
-
-    private static Stream<EntryData> flatten(EntryData entry) {
-
-        return Stream.concat(Stream.of(entry),
-                             entry.getChildrenList()
-                                     .stream()
-                                     .flatMap(DocumentStructureWrapper::flatten));
-    }
-
-}

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

@@ -10,7 +10,6 @@ import lombok.Data;
 import lombok.NoArgsConstructor;
 import lombok.experimental.FieldDefaults;
 
-@Deprecated
 @Data
 @Builder
 @NoArgsConstructor

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

@@ -1,6 +1,5 @@
 package com.knecon.fforesight.service.layoutparser.internal.api.data.redaction;
 
-@Deprecated
 public enum LayoutEngine {
     ALGORITHM,
     AI,

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

@@ -1,193 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.data.redaction;
-
-// Generated by the protocol buffer compiler.  DO NOT EDIT!
-// NO CHECKED-IN PROTOBUF GENCODE
-// source: LayoutEngine.proto
-// Protobuf Java Version: 4.27.1
-@SuppressWarnings("all")
-public final class LayoutEngineProto {
-
-    private LayoutEngineProto() {}
-
-
-    static {
-        com.google.protobuf.RuntimeVersion.validateProtobufGencodeVersion(com.google.protobuf.RuntimeVersion.RuntimeDomain.PUBLIC,
-                /* major= */ 4,
-                /* minor= */ 27,
-                /* patch= */ 1,
-                /* suffix= */ "", LayoutEngineProto.class.getName());
-    }
-
-    public static void registerAllExtensions(com.google.protobuf.ExtensionRegistryLite registry) {
-
-    }
-
-
-    public static void registerAllExtensions(com.google.protobuf.ExtensionRegistry registry) {
-
-        registerAllExtensions((com.google.protobuf.ExtensionRegistryLite) registry);
-    }
-
-
-    /**
-     * Protobuf enum {@code LayoutEngine}
-     */
-    public enum LayoutEngine implements com.google.protobuf.ProtocolMessageEnum {
-        /**
-         * <code>ALGORITHM = 0;</code>
-         */
-        ALGORITHM(0),
-        /**
-         * <code>AI = 1;</code>
-         */
-        AI(1),
-        /**
-         * <code>OUTLINE = 2;</code>
-         */
-        OUTLINE(2),
-        UNRECOGNIZED(-1),
-        ;
-
-
-        static {
-            com.google.protobuf.RuntimeVersion.validateProtobufGencodeVersion(com.google.protobuf.RuntimeVersion.RuntimeDomain.PUBLIC,
-                    /* major= */ 4,
-                    /* minor= */ 27,
-                    /* patch= */ 1,
-                    /* suffix= */ "", LayoutEngine.class.getName());
-        }
-
-        /**
-         * <code>ALGORITHM = 0;</code>
-         */
-        public static final int ALGORITHM_VALUE = 0;
-        /**
-         * <code>AI = 1;</code>
-         */
-        public static final int AI_VALUE = 1;
-        /**
-         * <code>OUTLINE = 2;</code>
-         */
-        public static final int OUTLINE_VALUE = 2;
-
-
-        public final int getNumber() {
-
-            if (this == UNRECOGNIZED) {
-                throw new IllegalArgumentException("Can't get the number of an unknown enum value.");
-            }
-            return value;
-        }
-
-
-        /**
-         * @param value The numeric wire value of the corresponding enum entry.
-         * @return The enum associated with the given numeric wire value.
-         * @deprecated Use {@link #forNumber(int)} instead.
-         */
-        @Deprecated
-        public static LayoutEngine valueOf(int value) {
-
-            return forNumber(value);
-        }
-
-
-        /**
-         * @param value The numeric wire value of the corresponding enum entry.
-         * @return The enum associated with the given numeric wire value.
-         */
-        public static LayoutEngine forNumber(int value) {
-
-            switch (value) {
-                case 0:
-                    return ALGORITHM;
-                case 1:
-                    return AI;
-                case 2:
-                    return OUTLINE;
-                default:
-                    return null;
-            }
-        }
-
-
-        public static com.google.protobuf.Internal.EnumLiteMap<LayoutEngine> internalGetValueMap() {
-
-            return internalValueMap;
-        }
-
-
-        private static final com.google.protobuf.Internal.EnumLiteMap<LayoutEngine> internalValueMap = new com.google.protobuf.Internal.EnumLiteMap<LayoutEngine>() {
-            public LayoutEngine findValueByNumber(int number) {
-
-                return LayoutEngine.forNumber(number);
-            }
-        };
-
-
-        public final com.google.protobuf.Descriptors.EnumValueDescriptor getValueDescriptor() {
-
-            if (this == UNRECOGNIZED) {
-                throw new IllegalStateException("Can't get the descriptor of an unrecognized enum value.");
-            }
-            return getDescriptor().getValues()
-                    .get(ordinal());
-        }
-
-
-        public final com.google.protobuf.Descriptors.EnumDescriptor getDescriptorForType() {
-
-            return getDescriptor();
-        }
-
-
-        public static final com.google.protobuf.Descriptors.EnumDescriptor getDescriptor() {
-
-            return LayoutEngineProto.getDescriptor().getEnumTypes()
-                    .get(0);
-        }
-
-
-        private static final LayoutEngine[] VALUES = values();
-
-
-        public static LayoutEngine valueOf(com.google.protobuf.Descriptors.EnumValueDescriptor desc) {
-
-            if (desc.getType() != getDescriptor()) {
-                throw new IllegalArgumentException("EnumValueDescriptor is not for this type.");
-            }
-            if (desc.getIndex() == -1) {
-                return UNRECOGNIZED;
-            }
-            return VALUES[desc.getIndex()];
-        }
-
-
-        private final int value;
-
-
-        private LayoutEngine(int value) {
-
-            this.value = value;
-        }
-
-        // @@protoc_insertion_point(enum_scope:LayoutEngine)
-    }
-
-
-    public static com.google.protobuf.Descriptors.FileDescriptor getDescriptor() {
-
-        return descriptor;
-    }
-
-
-    private static com.google.protobuf.Descriptors.FileDescriptor descriptor;
-
-    static {
-        String[] descriptorData = {"\n\022LayoutEngine.proto*2\n\014LayoutEngine\022\r\n\t" + "ALGORITHM\020\000\022\006\n\002AI\020\001\022\013\n\007OUTLINE\020\002b\006proto3"};
-        descriptor = com.google.protobuf.Descriptors.FileDescriptor.internalBuildGeneratedFileFrom(descriptorData, new com.google.protobuf.Descriptors.FileDescriptor[]{});
-        descriptor.resolveAllFeaturesImmutable();
-    }
-
-    // @@protoc_insertion_point(outer_class_scope)
-}

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

@@ -3,7 +3,6 @@ package com.knecon.fforesight.service.layoutparser.internal.api.data.redaction;
 import java.io.Serializable;
 import java.util.Locale;
 
-@Deprecated
 public enum NodeType implements Serializable {
     DOCUMENT,
     SECTION,

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

@@ -1,274 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.internal.api.data.redaction;
-
-import java.util.Locale;
-// Generated by the protocol buffer compiler.  DO NOT EDIT!
-// NO CHECKED-IN PROTOBUF GENCODE
-// source: NodeType.proto
-// Protobuf Java Version: 4.27.1
-@SuppressWarnings("all")
-public final class NodeTypeProto {
-
-    private NodeTypeProto() {}
-
-
-    static {
-        com.google.protobuf.RuntimeVersion.validateProtobufGencodeVersion(com.google.protobuf.RuntimeVersion.RuntimeDomain.PUBLIC,
-                /* major= */ 4,
-                /* minor= */ 27,
-                /* patch= */ 1,
-                /* suffix= */ "", NodeTypeProto.class.getName());
-    }
-
-    public static void registerAllExtensions(com.google.protobuf.ExtensionRegistryLite registry) {
-
-    }
-
-
-    public static void registerAllExtensions(com.google.protobuf.ExtensionRegistry registry) {
-
-        registerAllExtensions((com.google.protobuf.ExtensionRegistryLite) registry);
-    }
-
-
-    /**
-     * Protobuf enum {@code NodeType}
-     */
-    public enum NodeType implements com.google.protobuf.ProtocolMessageEnum {
-        /**
-         * <code>DOCUMENT = 0;</code>
-         */
-        DOCUMENT(0),
-        /**
-         * <code>SECTION = 1;</code>
-         */
-        SECTION(1),
-        /**
-         * <code>SUPER_SECTION = 2;</code>
-         */
-        SUPER_SECTION(2),
-        /**
-         * <code>HEADLINE = 3;</code>
-         */
-        HEADLINE(3),
-        /**
-         * <code>PARAGRAPH = 4;</code>
-         */
-        PARAGRAPH(4),
-        /**
-         * <code>TABLE = 5;</code>
-         */
-        TABLE(5),
-        /**
-         * <code>TABLE_CELL = 6;</code>
-         */
-        TABLE_CELL(6),
-        /**
-         * <code>IMAGE = 7;</code>
-         */
-        IMAGE(7),
-        /**
-         * <code>HEADER = 8;</code>
-         */
-        HEADER(8),
-        /**
-         * <code>FOOTER = 9;</code>
-         */
-        FOOTER(9),
-        UNRECOGNIZED(-1),
-        ;
-
-
-        public String toString() {
-
-            return this.name().charAt(0) + this.name().substring(1).toLowerCase(Locale.ROOT);
-        }
-
-
-        static {
-            com.google.protobuf.RuntimeVersion.validateProtobufGencodeVersion(com.google.protobuf.RuntimeVersion.RuntimeDomain.PUBLIC,
-                    /* major= */ 4,
-                    /* minor= */ 27,
-                    /* patch= */ 1,
-                    /* suffix= */ "", NodeType.class.getName());
-        }
-
-        /**
-         * <code>DOCUMENT = 0;</code>
-         */
-        public static final int DOCUMENT_VALUE = 0;
-        /**
-         * <code>SECTION = 1;</code>
-         */
-        public static final int SECTION_VALUE = 1;
-        /**
-         * <code>SUPER_SECTION = 2;</code>
-         */
-        public static final int SUPER_SECTION_VALUE = 2;
-        /**
-         * <code>HEADLINE = 3;</code>
-         */
-        public static final int HEADLINE_VALUE = 3;
-        /**
-         * <code>PARAGRAPH = 4;</code>
-         */
-        public static final int PARAGRAPH_VALUE = 4;
-        /**
-         * <code>TABLE = 5;</code>
-         */
-        public static final int TABLE_VALUE = 5;
-        /**
-         * <code>TABLE_CELL = 6;</code>
-         */
-        public static final int TABLE_CELL_VALUE = 6;
-        /**
-         * <code>IMAGE = 7;</code>
-         */
-        public static final int IMAGE_VALUE = 7;
-        /**
-         * <code>HEADER = 8;</code>
-         */
-        public static final int HEADER_VALUE = 8;
-        /**
-         * <code>FOOTER = 9;</code>
-         */
-        public static final int FOOTER_VALUE = 9;
-
-
-        public final int getNumber() {
-
-            if (this == UNRECOGNIZED) {
-                throw new IllegalArgumentException("Can't get the number of an unknown enum value.");
-            }
-            return value;
-        }
-
-
-        /**
-         * @param value The numeric wire value of the corresponding enum entry.
-         * @return The enum associated with the given numeric wire value.
-         * @deprecated Use {@link #forNumber(int)} instead.
-         */
-        @Deprecated
-        public static NodeType valueOf(int value) {
-
-            return forNumber(value);
-        }
-
-
-        /**
-         * @param value The numeric wire value of the corresponding enum entry.
-         * @return The enum associated with the given numeric wire value.
-         */
-        public static NodeType forNumber(int value) {
-
-            switch (value) {
-                case 0:
-                    return DOCUMENT;
-                case 1:
-                    return SECTION;
-                case 2:
-                    return SUPER_SECTION;
-                case 3:
-                    return HEADLINE;
-                case 4:
-                    return PARAGRAPH;
-                case 5:
-                    return TABLE;
-                case 6:
-                    return TABLE_CELL;
-                case 7:
-                    return IMAGE;
-                case 8:
-                    return HEADER;
-                case 9:
-                    return FOOTER;
-                default:
-                    return null;
-            }
-        }
-
-
-        public static com.google.protobuf.Internal.EnumLiteMap<NodeType> internalGetValueMap() {
-
-            return internalValueMap;
-        }
-
-
-        private static final com.google.protobuf.Internal.EnumLiteMap<NodeType> internalValueMap = new com.google.protobuf.Internal.EnumLiteMap<NodeType>() {
-            public NodeType findValueByNumber(int number) {
-
-                return NodeType.forNumber(number);
-            }
-        };
-
-
-        public final com.google.protobuf.Descriptors.EnumValueDescriptor getValueDescriptor() {
-
-            if (this == UNRECOGNIZED) {
-                throw new IllegalStateException("Can't get the descriptor of an unrecognized enum value.");
-            }
-            return getDescriptor().getValues()
-                    .get(ordinal());
-        }
-
-
-        public final com.google.protobuf.Descriptors.EnumDescriptor getDescriptorForType() {
-
-            return getDescriptor();
-        }
-
-
-        public static final com.google.protobuf.Descriptors.EnumDescriptor getDescriptor() {
-
-            return NodeTypeProto.getDescriptor().getEnumTypes()
-                    .get(0);
-        }
-
-
-        private static final NodeType[] VALUES = values();
-
-
-        public static NodeType valueOf(com.google.protobuf.Descriptors.EnumValueDescriptor desc) {
-
-            if (desc.getType() != getDescriptor()) {
-                throw new IllegalArgumentException("EnumValueDescriptor is not for this type.");
-            }
-            if (desc.getIndex() == -1) {
-                return UNRECOGNIZED;
-            }
-            return VALUES[desc.getIndex()];
-        }
-
-
-        private final int value;
-
-
-        private NodeType(int value) {
-
-            this.value = value;
-        }
-
-        // @@protoc_insertion_point(enum_scope:NodeType)
-    }
-
-
-    public static com.google.protobuf.Descriptors.FileDescriptor getDescriptor() {
-
-        return descriptor;
-    }
-
-
-    private static com.google.protobuf.Descriptors.FileDescriptor descriptor;
-
-    static {
-        String[] descriptorData = {"\n\016NodeType.proto*\223\001\n\010NodeType\022\014\n\010DOCUMEN"
-                                   + "T\020\000\022\013\n\007SECTION\020\001\022\021\n\rSUPER_SECTION\020\002\022\014\n\010H"
-                                   + "EADLINE\020\003\022\r\n\tPARAGRAPH\020\004\022\t\n\005TABLE\020\005\022\016\n\nT"
-                                   + "ABLE_CELL\020\006\022\t\n\005IMAGE\020\007\022\n\n\006HEADER\020\010\022\n\n\006FO"
-                                   + "OTER\020\tb\006proto3"};
-        descriptor = com.google.protobuf.Descriptors.FileDescriptor.internalBuildGeneratedFileFrom(descriptorData, new com.google.protobuf.Descriptors.FileDescriptor[]{});
-        descriptor.resolveAllFeaturesImmutable();
-    }
-
-    // @@protoc_insertion_point(outer_class_scope)
-}

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

@@ -19,16 +19,6 @@ public class SimplifiedText {
     @Schema(description = "Number of pages in the entire document.")
     private int numberOfPages;
     @Schema(description = "A List of simplified Sections, which contains almost exclusively text.")
-    @Builder.Default
     private List<SimplifiedSectionText> sectionTexts = new ArrayList<>();
-    @Schema(description = "A list of the main section numbers ")
-    @Builder.Default
-    private List<String> mainSectionNumbers = new ArrayList<>();
-    @Schema(description = "A list of the header section numbers ")
-    @Builder.Default
-    private List<String> headerSectionNumbers = new ArrayList<>();
-    @Schema(description = "A list of the footer section numbers ")
-    @Builder.Default
-    private List<String> footerSectionNumbers = new ArrayList<>();
 
 }

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

@@ -1,7 +1,5 @@
 package com.knecon.fforesight.service.layoutparser.internal.api.data.taas;
 
-import java.util.List;
-
 import io.swagger.v3.oas.annotations.media.Schema;
 import lombok.AllArgsConstructor;
 import lombok.Builder;
@@ -15,8 +13,6 @@ public class StructureObject {
 
     @Schema(description = "The ID of this StructureObject.")
     Integer structureObjectNumber;
-    @Schema(description = "The Tree ID of this StructureObject.")
-    List<Integer> treeId;
     @Schema(description = "This value indicates the start of the string offsets in this Object, with respect to the reading order.")
     int page;
     @Schema(description = "This stringOffset indicates the start of the string offsets in this Object, with respect to the reading order of the entire document. It is equal to the previous' StructureObject stringOffset + its length.")

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

@@ -2,9 +2,6 @@ package com.knecon.fforesight.service.layoutparser.internal.api.queue;
 
 public class LayoutParsingQueueNames {
 
-    public static final String LAYOUT_PARSING_REQUEST_QUEUE_PREFIX = "layout_parsing_request";
-    public static final String LAYOUT_PARSING_REQUEST_EXCHANGE = "layout_parsing_request_exchange";
-    public static final String LAYOUT_PARSING_RESPONSE_QUEUE_PREFIX = "layout_parsing_response";
-    public static final String LAYOUT_PARSING_RESPONSE_EXCHANGE = "layout_parsing_response_exchange";
-    public static final String LAYOUT_PARSING_DLQ = "layout_parsing_error";
+    public static final String LAYOUT_PARSING_REQUEST_QUEUE = "layout_parsing_request_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

@@ -10,37 +10,24 @@ import lombok.NonNull;
 @Builder
 @Schema(description = "Object containing all storage paths the service needs to know.")
 public record LayoutParsingRequest(
-        @Schema(description = "Enum specifying the type of layout parsing to be performed.", allowableValues = "{RedactManager, DocuMine, TAAS}")//
         @NonNull LayoutParsingType layoutParsingType,
 
-        @Schema(description = "General purpose identifiers. They are not changed by the service at all and are returned as is in the response queue.")//
         Map<String, String> identifier,
 
-        @Schema(description = "Path to the original PDF file.")//
-        @NonNull String originFileStorageId,//
+        @NonNull String originFileStorageId,
 
-        @Schema(description = "Optional Path to the table extraction file.")//
-        Optional<String> tablesFileStorageId,//
-        @Schema(description = "Optional Path to the image classification file.")//
-        Optional<String> imagesFileStorageId,//
+        Optional<String> tablesFileStorageId,
+        Optional<String> imagesFileStorageId,
 
-        @Schema(description = "Optional Path to the the visual layout parsing service file") Optional<String> visualLayoutParsingFileId,//
+        Optional<String> visualLayoutParsingFileId,
 
-        @Schema(description = "Path where the Document Structure File will be stored.")//
-        @NonNull String structureFileStorageId,//
-        @Schema(description = "Path where the Research Data File will be stored.")//
-        String researchDocumentStorageId,//
-        @Schema(description = "Path where the Document Text File will be stored.")//
-        @NonNull String textBlockFileStorageId,//
-        @Schema(description = "Path where the Document Positions File will be stored.")//
-        @NonNull String positionBlockFileStorageId,//
-        @Schema(description = "Path where the Document Pages File will be stored.")//
-        @NonNull String pageFileStorageId,//
-        @Schema(description = "Path where the Document Markdown File will be stored.")//
-        Optional<String> documentMarkdownFileStorageId,//
-        @Schema(description = "Path where the Simplified Text File will be stored.")//
-        @NonNull String simplifiedTextStorageId,//
-        @Schema(description = "Path where the Viewer Document PDF will be stored.")//
+        @NonNull String structureFileStorageId,
+        String researchDocumentStorageId,
+        String markdownDocumentStorageId,
+        @NonNull String textBlockFileStorageId,
+        @NonNull String positionBlockFileStorageId,
+        @NonNull String pageFileStorageId,
+        @NonNull String simplifiedTextStorageId,
         @NonNull String viewerDocumentStorageId
 ) {
 

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

@@ -8,5 +8,6 @@ public enum LayoutParsingType {
     DOCUMINE,
     DOCUMINE_OLD,
     CLARIFYND,
-    CLARIFYND_PARAGRAPH_DEBUG
+    CLARIFYND_PARAGRAPH_DEBUG,
+    MARKDOWN
 }

layoutparser-service/layoutparser-service-internal-api/src/main/resources/DocumentPage.proto

@@ -1,21 +0,0 @@
-syntax = "proto3";
-
-message AllDocumentPages {
-
-  repeated DocumentPage documentPages = 1;
-}
-
-message DocumentPage {
-  // The page number, starting with 1.
-  int32 number = 1;
-
-  // The page height in PDF user units.
-  int32 height = 2;
-
-  // The page width in PDF user units.
-  int32 width = 3;
-
-  // The page rotation as specified by the PDF.
-  int32 rotation = 4;
-}
-

layoutparser-service/layoutparser-service-internal-api/src/main/resources/DocumentPositionData.proto

@@ -1,25 +0,0 @@
-syntax = "proto3";
-
-message AllDocumentPositionData {
-
-  repeated DocumentPositionData documentPositionData = 1;
-}
-
-message DocumentPositionData {
-  // Identifier of the text block.
-  int64 id = 1;
-
-  // For each string coordinate in the search text of the text block, the array contains an entry relating the string coordinate to the position coordinate.
-  // This is required due to the text and position coordinates not being equal.
-  repeated int32 stringIdxToPositionIdx = 2;
-
-  // The bounding box for each glyph as a rectangle. This matrix is of size (n,4), where n is the number of glyphs in the text block.
-  // The second dimension specifies the rectangle with the value x, y, width, height, with x, y specifying the lower left corner.
-  // In order to access this information, the stringIdxToPositionIdx array must be used to transform the coordinates.
-  repeated Position positions = 3;
-
-  // Definition of a BoundingBox that contains x, y, width, and height.
-  message Position {
-    repeated float value = 1;
-  }
-}

layoutparser-service/layoutparser-service-internal-api/src/main/resources/DocumentStructure.proto

@@ -1,8 +0,0 @@
-syntax = "proto3";
-
-import "EntryData.proto";
-
-message DocumentStructure {
-  // The root EntryData represents the Document.
-  EntryData root = 1;
-}

layoutparser-service/layoutparser-service-internal-api/src/main/resources/DocumentTextData.proto

@@ -1,29 +0,0 @@
-syntax = "proto3";
-
-message AllDocumentTextData {
-
-  repeated DocumentTextData documentTextData = 1;
-}
-
-message DocumentTextData {
-  // Identifier of the text block.
-  int64 id = 1;
-
-  // The page the text block occurs on.
-  int64 page = 2;
-
-  // The text of the text block.
-  string searchText = 3;
-
-  // Each text block is assigned a number on a page, starting from 0.
-  int32 numberOnPage = 4;
-
-  // The text blocks are ordered, this number represents the start of the text block as a string offset.
-  int32 start = 5;
-
-  // The text blocks are ordered, this number represents the end of the text block as a string offset.
-  int32 end = 6;
-
-  // The line breaks in the text of this semantic node in string offsets. They are exclusive end. At the end of each semantic node there is an implicit linebreak.
-  repeated int32 lineBreaks = 7;
-}

layoutparser-service/layoutparser-service-internal-api/src/main/resources/EntryData.proto

@@ -1,27 +0,0 @@
-syntax = "proto3";
-
-import "LayoutEngine.proto";
-import "NodeType.proto";
-
-message EntryData {
-  // Type of the semantic node.
-  NodeType type = 1;
-
-  // Specifies the position in the parsed tree structure.
-  repeated int32 treeId = 2;
-
-  // Specifies the text block IDs associated with this semantic node.
-  repeated int64 atomicBlockIds = 3;
-
-  // Specifies the pages this semantic node appears on.
-  repeated int64 pageNumbers = 4;
-
-  // Some semantic nodes have additional information, this information is stored in this Map.
-  map<string, string> properties = 5;
-
-  // All child Entries of this Entry.
-  repeated EntryData children = 6;
-
-  // Describes the origin of the semantic node.
-  repeated LayoutEngine engines = 7;
-}

layoutparser-service/layoutparser-service-internal-api/src/main/resources/LayoutEngine.proto

@@ -1,7 +0,0 @@
-syntax = "proto3";
-
-enum LayoutEngine {
-  ALGORITHM = 0;
-  AI = 1;
-  OUTLINE = 2;
-}

layoutparser-service/layoutparser-service-internal-api/src/main/resources/NodeType.proto

@@ -1,14 +0,0 @@
-syntax = "proto3";
-
-enum NodeType {
-  DOCUMENT = 0;
-  SECTION = 1;
-  SUPER_SECTION = 2;
-  HEADLINE = 3;
-  PARAGRAPH = 4;
-  TABLE = 5;
-  TABLE_CELL = 6;
-  IMAGE = 7;
-  HEADER = 8;
-  FOOTER = 9;
-}

layoutparser-service/layoutparser-service-processor/build.gradle.kts

@@ -12,14 +12,12 @@ dependencies {
     implementation(project(":layoutparser-service-internal-api"))
     implementation(project(":viewer-doc-processor"))
 
-    implementation("com.iqser.red.service:persistence-service-shared-api-v1:2.564.0-RED9010.0") {
+    implementation("com.iqser.red.service:persistence-service-shared-api-v1:2.144.0") {
         exclude("org.springframework.boot", "spring-boot-starter-security")
         exclude("org.springframework.boot", "spring-boot-starter-validation")
     }
-    implementation("com.knecon.fforesight:tenant-commons:0.30.0") {
-        exclude("com.iqser.red.commons", "storage-commons")
-    }
-    implementation("com.iqser.red.commons:storage-commons:2.50.0")
+    implementation("com.knecon.fforesight:tenant-commons:0.21.0")
+    implementation("com.iqser.red.commons:storage-commons:2.45.0")
 
     implementation("org.apache.pdfbox:pdfbox:${pdfBoxVersion}")
     implementation("org.apache.pdfbox:pdfbox-tools:${pdfBoxVersion}")
@@ -27,14 +25,11 @@ dependencies {
     implementation("com.fasterxml.jackson.datatype:jackson-datatype-jsr310:${jacksonVersion}")
     implementation("org.springframework.boot:spring-boot-starter-web:3.1.3")
     implementation("org.jgrapht:jgrapht-core:1.5.2")
-    implementation("org.apache.pdfbox:jbig2-imageio:3.0.4")
-    implementation("com.github.jai-imageio:jai-imageio-core:1.4.0")
-    implementation("com.github.jai-imageio:jai-imageio-jpeg2000:1.4.0")
     implementation("org.tinspin:tinspin-indexes:2.1.3")
     implementation("org.commonmark:commonmark:0.22.0")
     implementation("org.commonmark:commonmark-ext-gfm-tables:0.22.0")
-    implementation("com.pdftron:PDFNet:10.11.0")
-    implementation("org.apache.commons:commons-text:1.12.0")
+    implementation("com.didalgo:gpt3-tokenizer:0.1.8")
 
-    implementation("com.google.protobuf:protobuf-java-util:4.27.1")
+    implementation("org.mapstruct:mapstruct:1.5.5.Final")
+    annotationProcessor("org.mapstruct:mapstruct-processor:1.5.5.Final")
 }

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

@@ -2,15 +2,14 @@ package com.knecon.fforesight.service.layoutparser.processor;
 
 import static java.lang.String.format;
 
-import java.awt.geom.AffineTransform;
+import java.awt.geom.Point2D;
 import java.awt.geom.Rectangle2D;
+import java.io.ByteArrayInputStream;
 import java.io.File;
 import java.io.IOException;
 import java.nio.file.Files;
 import java.util.ArrayList;
-import java.util.Collection;
 import java.util.HashMap;
-import java.util.HashSet;
 import java.util.List;
 import java.util.Map;
 import java.util.concurrent.atomic.AtomicReference;
@@ -20,18 +19,15 @@ import org.apache.pdfbox.pdmodel.PDDocument;
 import org.apache.pdfbox.pdmodel.PDPage;
 import org.apache.pdfbox.pdmodel.common.PDRectangle;
 import org.apache.pdfbox.pdmodel.documentinterchange.markedcontent.PDMarkedContent;
+import org.commonmark.ext.gfm.tables.TablesExtension;
+import org.commonmark.renderer.markdown.MarkdownRenderer;
 import org.springframework.stereotype.Service;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeTypeProto.NodeType;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
 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.internal.api.queue.LayoutParsingType;
-import com.knecon.fforesight.service.layoutparser.processor.docstrum.model.Character;
-import com.knecon.fforesight.service.layoutparser.processor.model.table.Ruling;
-import com.knecon.fforesight.service.layoutparser.processor.model.text.RedTextPosition;
-import com.knecon.fforesight.service.layoutparser.processor.model.text.TextDirection;
-import com.knecon.fforesight.service.layoutparser.processor.services.classification.ClassificationService;
-import com.knecon.fforesight.service.layoutparser.processor.services.mapper.MarkdownMapper;
+import com.knecon.fforesight.service.layoutparser.processor.markdown.DocumentDataParser;
 import com.knecon.fforesight.service.layoutparser.processor.model.AbstractPageBlock;
 import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationDocument;
 import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationPage;
@@ -46,7 +42,7 @@ import com.knecon.fforesight.service.layoutparser.processor.model.outline.TableO
 import com.knecon.fforesight.service.layoutparser.processor.model.table.Cell;
 import com.knecon.fforesight.service.layoutparser.processor.model.table.CleanRulings;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
-import com.knecon.fforesight.service.layoutparser.processor.model.text.Word;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPositionSequence;
 import com.knecon.fforesight.service.layoutparser.processor.python_api.adapter.CvTableParsingAdapter;
 import com.knecon.fforesight.service.layoutparser.processor.python_api.adapter.ImageServiceResponseAdapter;
 import com.knecon.fforesight.service.layoutparser.processor.python_api.adapter.VisualLayoutParsingAdapter;
@@ -54,6 +50,7 @@ import com.knecon.fforesight.service.layoutparser.processor.python_api.model.ima
 import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableCells;
 import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableServiceResponse;
 import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.VisualLayoutParsingResponse;
+import com.knecon.fforesight.service.layoutparser.processor.services.BodyTextFrameService;
 import com.knecon.fforesight.service.layoutparser.processor.services.RulingCleaningService;
 import com.knecon.fforesight.service.layoutparser.processor.services.SectionsBuilderService;
 import com.knecon.fforesight.service.layoutparser.processor.services.SimplifiedSectionTextService;
@@ -63,6 +60,9 @@ import com.knecon.fforesight.service.layoutparser.processor.services.blockificat
 import com.knecon.fforesight.service.layoutparser.processor.services.blockification.DocstrumBlockificationService;
 import com.knecon.fforesight.service.layoutparser.processor.services.blockification.DocuMineBlockificationService;
 import com.knecon.fforesight.service.layoutparser.processor.services.blockification.RedactManagerBlockificationService;
+import com.knecon.fforesight.service.layoutparser.processor.services.classification.ClarifyndClassificationService;
+import com.knecon.fforesight.service.layoutparser.processor.services.classification.DocuMineClassificationService;
+import com.knecon.fforesight.service.layoutparser.processor.services.classification.RedactManagerClassificationService;
 import com.knecon.fforesight.service.layoutparser.processor.services.factory.DocumentGraphFactory;
 import com.knecon.fforesight.service.layoutparser.processor.services.graphics.Box;
 import com.knecon.fforesight.service.layoutparser.processor.services.graphics.GraphicExtractorService;
@@ -72,7 +72,6 @@ import com.knecon.fforesight.service.layoutparser.processor.services.parsing.PDF
 import com.knecon.fforesight.service.layoutparser.processor.services.visualization.LayoutGridService;
 import com.knecon.fforesight.service.layoutparser.processor.utils.MarkedContentUtils;
 import com.knecon.fforesight.service.layoutparser.processor.utils.PageInformation;
-import com.knecon.fforesight.service.layoutparser.processor.utils.TextPositionOperations;
 
 import io.micrometer.observation.Observation;
 import io.micrometer.observation.ObservationRegistry;
@@ -94,7 +93,10 @@ public class LayoutParsingPipeline {
     CvTableParsingAdapter cvTableParsingAdapter;
     LayoutParsingStorageService layoutParsingStorageService;
     SectionsBuilderService sectionsBuilderService;
+    RedactManagerClassificationService redactManagerClassificationService;
+    DocuMineClassificationService docuMineClassificationService;
     SimplifiedSectionTextService simplifiedSectionTextService;
+    BodyTextFrameService bodyTextFrameService;
     RulingCleaningService rulingCleaningService;
     TableExtractionService tableExtractionService;
     DocuMineBlockificationService docuMineBlockificationService;
@@ -104,12 +106,12 @@ public class LayoutParsingPipeline {
     LayoutGridService layoutGridService;
     ObservationRegistry observationRegistry;
     VisualLayoutParsingAdapter visualLayoutParsingAdapter;
+    ClarifyndClassificationService clarifyndClassificationService;
     GraphicExtractorService graphicExtractorService;
     OutlineExtractorService outlineExtractorService;
     OutlineValidationService outlineValidationService;
     TOCEnrichmentService tocEnrichmentService;
     LayoutparserSettings settings;
-    ClassificationService classificationService;
 
 
     public LayoutParsingFinishedEvent parseLayoutAndSaveFilesToStorage(LayoutParsingRequest layoutParsingRequest) throws IOException {
@@ -118,14 +120,23 @@ public class LayoutParsingPipeline {
         log.info("Starting layout parsing for {}", layoutParsingRequest.identifier());
 
         File originFile = layoutParsingStorageService.getOriginFile(layoutParsingRequest.originFileStorageId());
-        File viewerDocumentFile = layoutParsingStorageService.getViewerDocFile(layoutParsingRequest.viewerDocumentStorageId()).orElse(originFile);
+//        File viewerDocumentFile = layoutParsingStorageService.getViewerDocFile(layoutParsingRequest.viewerDocumentStorageId()).orElse(originFile);
+        File viewerDocumentFile = originFile;
 
-        VisualLayoutParsingResponse visualLayoutParsingResponse = layoutParsingRequest.visualLayoutParsingFileId()
-                .map(layoutParsingStorageService::getVisualLayoutParsingFile).orElse(new VisualLayoutParsingResponse());
-        ImageServiceResponse imageServiceResponse = layoutParsingRequest.imagesFileStorageId()
-                .map(layoutParsingStorageService::getImagesFile).orElse(new ImageServiceResponse());
-        TableServiceResponse tableServiceResponse = layoutParsingRequest.tablesFileStorageId()
-                .map(layoutParsingStorageService::getTablesFile).orElse(new TableServiceResponse());
+        VisualLayoutParsingResponse visualLayoutParsingResponse = new VisualLayoutParsingResponse();
+        if (layoutParsingRequest.visualLayoutParsingFileId().isPresent()) {
+            visualLayoutParsingResponse = layoutParsingStorageService.getVisualLayoutParsingFile(layoutParsingRequest.visualLayoutParsingFileId().get());
+        }
+
+        ImageServiceResponse imageServiceResponse = new ImageServiceResponse();
+        if (layoutParsingRequest.imagesFileStorageId().isPresent()) {
+            imageServiceResponse = layoutParsingStorageService.getImagesFile(layoutParsingRequest.imagesFileStorageId().get());
+        }
+
+        TableServiceResponse tableServiceResponse = new TableServiceResponse();
+        if (layoutParsingRequest.tablesFileStorageId().isPresent()) {
+            tableServiceResponse = layoutParsingStorageService.getTablesFile(layoutParsingRequest.tablesFileStorageId().get());
+        }
 
         ClassificationDocument classificationDocument = parseLayout(settings.getLayoutParsingTypeOverride() == null //
                                                                             ? layoutParsingRequest.layoutParsingType() : settings.getLayoutParsingTypeOverride(),
@@ -142,14 +153,11 @@ public class LayoutParsingPipeline {
 
         log.info("Creating viewer document for {}", layoutParsingRequest.identifier());
 
-        layoutGridService.addLayoutGrid(viewerDocumentFile, documentGraph, viewerDocumentFile, false);
+        layoutGridService.addLayoutGrid(viewerDocumentFile, documentGraph, viewerDocumentFile, false, layoutParsingRequest.visualLayoutParsingFileId().isPresent());
 
         log.info("Storing resulting files for {}", layoutParsingRequest.identifier());
 
         layoutParsingStorageService.storeDocumentData(layoutParsingRequest, DocumentDataMapper.toDocumentData(documentGraph));
-        if (layoutParsingRequest.documentMarkdownFileStorageId().isPresent()) {
-            layoutParsingStorageService.storeMarkdownFile(layoutParsingRequest.documentMarkdownFileStorageId().get(), new MarkdownMapper().toMarkdownContent(documentGraph));
-        }
         layoutParsingStorageService.storeSimplifiedText(layoutParsingRequest, simplifiedSectionTextService.toSimplifiedText(documentGraph));
         layoutParsingStorageService.storeViewerDocument(layoutParsingRequest, viewerDocumentFile);
 
@@ -159,10 +167,20 @@ public class LayoutParsingPipeline {
             layoutParsingStorageService.storeResearchDocumentData(layoutParsingRequest, researchDocumentData);
         }
 
-        if (!viewerDocumentFile.equals(originFile)) {
-            assert !viewerDocumentFile.exists() || viewerDocumentFile.delete();
+        if (layoutParsingRequest.markdownDocumentStorageId() != null) {
+            log.info("Rendering document data as markdown for {}", layoutParsingRequest.identifier());
+            var markdownDocument = DocumentDataParser.parse(documentGraph.streamAllSubNodes());
+            MarkdownRenderer renderer = MarkdownRenderer.builder().extensions(List.of(TablesExtension.create())).build();
+            String markdown = renderer.render(markdownDocument);
+            try (var in = new ByteArrayInputStream(markdown.getBytes())) {
+                layoutParsingStorageService.storeObject(layoutParsingRequest.markdownDocumentStorageId(), in);
+            }
         }
-        assert !originFile.exists() || originFile.delete();
+
+        if (!viewerDocumentFile.equals(originFile)) {
+            viewerDocumentFile.delete();
+        }
+        originFile.delete();
 
         return LayoutParsingFinishedEvent.builder()
                 .identifier(layoutParsingRequest.identifier())
@@ -234,16 +252,19 @@ public class LayoutParsingPipeline {
         Map<Integer, List<TableCells>> pdfTableCells = cvTableParsingAdapter.buildCvParsedTablesPerPage(tableServiceResponse);
         Map<Integer, List<ClassifiedImage>> pdfImages = imageServiceResponseAdapter.buildClassifiedImagesPerPage(imageServiceResponse);
         Map<Integer, List<ClassifiedImage>> signatures = visualLayoutParsingAdapter.buildExtractedSignaturesPerPage(visualLayoutParsingResponse);
-
         ClassificationDocument classificationDocument = new ClassificationDocument();
 
         if (settings.isDebug() || identifier.containsKey("debug")) {
-            classificationDocument.getLayoutDebugLayer().setActive(true);
+            classificationDocument.getVisualizations().setActive(true);
         }
 
         List<ClassificationPage> classificationPages = new ArrayList<>();
+        OutlineObject lastProcessedOutlineObject = null;
 
-        classificationDocument.setOutlineObjectTree(outlineExtractorService.getOutlineObjectTree(originDocument));
+        // parsing the structure elements could be useful as well
+        if (layoutParsingType != LayoutParsingType.REDACT_MANAGER_OLD && layoutParsingType != LayoutParsingType.DOCUMINE_OLD) {
+            classificationDocument.setOutlineObjectTree(outlineExtractorService.getOutlineObjectTree(originDocument));
+        }
 
         long pageCount = originDocument.getNumberOfPages();
 
@@ -267,55 +288,70 @@ public class LayoutParsingPipeline {
             stripper.setStartPage(pageNumber);
             stripper.setEndPage(pageNumber);
             stripper.setPdpage(pdPage);
-            stripper.getText(originDocument);
-            List<Word> words = stripper.getWords();
-
-//            rotateDirAdjExactly(words, pdPage); // works really well for many highly rotated documents (e.g. VV-331340.pdf), but it decreases the headline performance by 1.3%, so I am leaving it out for now
-
             if (layoutParsingType.equals(LayoutParsingType.DOCUMINE_OLD)) {
-                var lines = TextPositionOperations.groupByLine(new HashSet<>(words));
-                classificationDocument.getLayoutDebugLayer().addLineVisualizationsFromNestedTextPosition(lines, pageNumber);
-                words = TextPositionOperations.sortWords(lines);
+                stripper.setSortByPosition(true);
             }
-            classificationDocument.getLayoutDebugLayer().addTextVisualizations(words, pageNumber);
+            stripper.getText(originDocument);
+            List<TextPositionSequence> words = stripper.getTextPositionSequences();
+            classificationDocument.getVisualizations().addTextVisualizations(words, pageNumber);
 
             PDRectangle pdr = pdPage.getMediaBox();
 
-            List<Ruling> rulings = stripper.getRulings();
-            classificationDocument.getLayoutDebugLayer().addRulingVisualization(rulings, pageNumber);
-            CleanRulings cleanRulings = rulingCleaningService.deduplicateAndStraightenRulings(pdfTableCells.get(pageNumber), rulings);
+            int rotation = pdPage.getRotation();
+            boolean isLandscape = pdr.getWidth() > pdr.getHeight() && (rotation == 0 || rotation == 180) || pdr.getHeight() > pdr.getWidth() && (rotation == 90 || rotation == 270);
+
+            PDRectangle cropbox = pdPage.getCropBox();
+            classificationDocument.getVisualizations().addRulingVisualization(stripper.getRulings(), pageNumber);
+            CleanRulings cleanRulings = rulingCleaningService.deduplicateAndStraightenRulings(pdfTableCells.get(pageNumber), stripper.getRulings());
 
             PageInformation pageInformation = PageInformation.fromPDPage(pageNumber, pdPage);
             List<Cell> emptyTableCells = TableExtractionService.findCells(cleanRulings.getHorizontals(), cleanRulings.getVerticals(), pageInformation);
-            classificationDocument.getLayoutDebugLayer().addCellVisualizations(emptyTableCells, pageNumber);
+            classificationDocument.getVisualizations().addCellVisualizations(emptyTableCells, pageNumber);
 
             TextRulingsClassifier.classifyUnderlinedAndStrikethroughText(words, cleanRulings);
 
-            List<Box> graphics = graphicExtractorService.extractPathElementGraphics(originDocument, pdPage, pageNumber, cleanRulings, stripper.getWords(), false);
+            List<Box> graphics = graphicExtractorService.extractPathElementGraphics(originDocument, pdPage, pageNumber, cleanRulings, stripper.getTextPositionSequences(),
+
+                                                                                    false);
 
             pdfImages.computeIfAbsent(pageNumber, x -> new ArrayList<>())
                     .addAll(graphics.stream()
-                                    .map(g -> new ClassifiedImage(new Rectangle2D.Double(g.x1, g.y1, g.width(), g.height()),
-                                                                  ImageType.GRAPHIC,
-                                                                  false,
-                                                                  stripper.getPageNumber(),
-                                                                  ""))
+                                    .map(g -> new ClassifiedImage(new Rectangle2D.Double(g.x1, g.y1, g.width(), g.height()), ImageType.GRAPHIC, false, stripper.getPageNumber()))
                                     .toList());
 
             ClassificationPage classificationPage = switch (layoutParsingType) {
-                case REDACT_MANAGER_OLD -> redactManagerBlockificationService.blockify(stripper.getWords(), cleanRulings, classificationDocument.getLayoutDebugLayer());
+                case REDACT_MANAGER_OLD ->
+                        redactManagerBlockificationService.blockify(stripper.getTextPositionSequences(), cleanRulings, classificationDocument.getVisualizations());
                 case DOCUMINE_OLD -> docuMineBlockificationService.blockify(words, cleanRulings);
                 case DOCUMINE, REDACT_MANAGER, REDACT_MANAGER_PARAGRAPH_DEBUG, REDACT_MANAGER_WITHOUT_DUPLICATE_PARAGRAPH ->
-                        docstrumBlockificationService.blockify(words, cleanRulings, true, classificationDocument.getLayoutDebugLayer(), layoutParsingType);
+                        docstrumBlockificationService.blockify(words, cleanRulings, true, classificationDocument.getVisualizations(), layoutParsingType);
                 case CLARIFYND, CLARIFYND_PARAGRAPH_DEBUG ->
-                        docstrumBlockificationService.blockify(words, cleanRulings, false, classificationDocument.getLayoutDebugLayer(), layoutParsingType);
+                        docstrumBlockificationService.blockify(words, cleanRulings, false, classificationDocument.getVisualizations(), layoutParsingType);
+                default -> throw new IllegalArgumentException("Unexpected LayoutParsingType: " + layoutParsingType);
             };
 
-            updateClassificationPage(pdPage, pdr, classificationPage, cleanRulings, pageNumber, pageInformation);
+            classificationPage.setCleanRulings(cleanRulings);
+            classificationPage.setRotation(rotation);
+            classificationPage.setLandscape(isLandscape);
+            classificationPage.setPageNumber(pageNumber);
+            classificationPage.setPageWidth(cropbox.getWidth());
+            classificationPage.setPageHeight(cropbox.getHeight());
 
-            blockificationPostprocessingService.findHeadlinesFromOutline(classificationDocument, pageNumber, classificationPage, pageInformation);
+            if (layoutParsingType != LayoutParsingType.REDACT_MANAGER_OLD && layoutParsingType != LayoutParsingType.DOCUMINE_OLD) {
+                List<OutlineObject> outlineObjects = classificationDocument.getOutlineObjectTree().getOutlineObjectsPerPage().getOrDefault(pageNumber - 1, new ArrayList<>());
 
-            classificationDocument.getLayoutDebugLayer().addMarkedContentVisualizations(stripper.getMarkedContents(), pageNumber);
+                OutlineObject notFoundOutlineObject = null;
+                if (lastProcessedOutlineObject != null && !lastProcessedOutlineObject.isFound()) {
+                    lastProcessedOutlineObject.setPoint(new Point2D.Float(0, cropbox.getHeight()));
+                    notFoundOutlineObject = lastProcessedOutlineObject;
+                }
+                if (!outlineObjects.isEmpty()) {
+                    classificationPage.setOutlineObjects(outlineObjects);
+                    lastProcessedOutlineObject = blockificationPostprocessingService.sanitizeOutlineBlocks(classificationPage, notFoundOutlineObject);
+                }
+            }
+
+            classificationDocument.getVisualizations().addMarkedContentVisualizations(stripper.getMarkedContents(), pageNumber);
             // MarkedContent needs to be converted at this point, otherwise it leads to GC Problems in Pdfbox.
             classificationPage.setMarkedContentBboxPerType(convertMarkedContents(stripper.getMarkedContents()));
 
@@ -343,9 +379,27 @@ public class LayoutParsingPipeline {
 
         originDocument.close();
 
-        classificationService.classify(classificationDocument, layoutParsingType, identifier);
+        log.info("Calculating BodyTextFrame for {}", identifier);
+        bodyTextFrameService.setBodyTextFrames(classificationDocument, layoutParsingType);
+        for (ClassificationPage page : classificationDocument.getPages()) {
+            classificationDocument.getVisualizations().addCleanRulingVisualization(page.getCleanRulings(), page.getPageNumber());
+        }
+        log.info("Classify TextBlocks for {}", identifier);
+        switch (layoutParsingType) {
+            case REDACT_MANAGER, REDACT_MANAGER_PARAGRAPH_DEBUG, REDACT_MANAGER_OLD, CLARIFYND_PARAGRAPH_DEBUG, REDACT_MANAGER_WITHOUT_DUPLICATE_PARAGRAPH ->
+                    redactManagerClassificationService.classifyDocument(classificationDocument);
+            case DOCUMINE_OLD, DOCUMINE -> docuMineClassificationService.classifyDocument(classificationDocument);
+            case CLARIFYND -> clarifyndClassificationService.classifyDocument(classificationDocument);
+        }
 
-        TableOfContents tableOfContents = outlineValidationService.createToC(classificationDocument);
+        List<TextPageBlock> headlines = classificationDocument.getPages()
+                .stream()
+                .flatMap(classificationPage -> classificationPage.getTextBlocks()
+                        .stream()
+                        .filter(tb -> tb instanceof TextPageBlock && tb.getClassification() != null && tb.getClassification().isHeadline())
+                        .map(tb -> (TextPageBlock) tb))
+                .toList();
+        TableOfContents tableOfContents = outlineValidationService.createToC(headlines);
         classificationDocument.setTableOfContents(tableOfContents);
 
         log.info("Building Sections for {}", identifier);
@@ -359,50 +413,6 @@ public class LayoutParsingPipeline {
     }
 
 
-    private static void updateClassificationPage(PDPage pdPage,
-                                                 PDRectangle pdr,
-                                                 ClassificationPage classificationPage,
-                                                 CleanRulings cleanRulings,
-                                                 int pageNumber,
-                                                 PageInformation pageInformation) {
-
-        int rotation = pdPage.getRotation();
-        boolean isLandscape = pdr.getWidth() > pdr.getHeight() && (rotation == 0 || rotation == 180) || pdr.getHeight() > pdr.getWidth() && (rotation == 90 || rotation == 270);
-        classificationPage.setCleanRulings(cleanRulings);
-        classificationPage.setRotation(rotation);
-        classificationPage.setLandscape(isLandscape);
-        classificationPage.setPageNumber(pageNumber);
-        classificationPage.setPageWidth((float) pageInformation.width());
-        classificationPage.setPageHeight((float) pageInformation.height());
-    }
-
-
-    private static void rotateDirAdjExactly(List<Word> words, PDPage pdPage) {
-
-        for (TextDirection dir : TextDirection.values()) {
-            double averageRotation = words.stream()
-                    .map(Word::getCharacters)
-                    .flatMap(Collection::stream)
-                    .map(Character::getTextPosition)
-                    .filter(pos -> pos.getDir().equals(dir))
-                    .mapToDouble(RedTextPosition::getExactDir).average().orElse(0);
-
-            if (averageRotation == 0) {
-                continue;
-            }
-
-            AffineTransform rotateInstance = AffineTransform.getRotateInstance(averageRotation, pdPage.getMediaBox().getWidth() / 2, pdPage.getMediaBox().getHeight() / 2);
-
-            for (Word word : words) {
-                if (!dir.equals(word.getDir())) {
-                    continue;
-                }
-                word.transform(rotateInstance);
-            }
-        }
-    }
-
-
     private void addNumberOfPagesToTrace(int numberOfPages, long size) {
 
         if (observationRegistry.getCurrentObservation() != null) {
@@ -444,10 +454,10 @@ public class LayoutParsingPipeline {
         // Collect all statistics for the classificationPage, except from blocks inside tables, as tables will always be added to BodyTextFrame.
         for (AbstractPageBlock textBlock : classificationPage.getTextBlocks()) {
             if (textBlock instanceof TextPageBlock) {
-                if (((TextPageBlock) textBlock).getWords() == null) {
+                if (((TextPageBlock) textBlock).getSequences() == null) {
                     continue;
                 }
-                for (Word word : ((TextPageBlock) textBlock).getWords()) {
+                for (TextPositionSequence word : ((TextPageBlock) textBlock).getSequences()) {
                     classificationPage.getTextHeightCounter().add(word.getTextHeight());
                     classificationPage.getFontCounter().add(word.getFont());
                     classificationPage.getFontSizeCounter().add(word.getFontSize());

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

@@ -5,7 +5,7 @@ import org.springframework.context.annotation.Bean;
 import org.springframework.context.annotation.ComponentScan;
 import org.springframework.context.annotation.Configuration;
 
-import com.knecon.fforesight.service.viewerdoc.service.PDFTronViewerDocumentService;
+import com.knecon.fforesight.service.viewerdoc.service.ViewerDocumentService;
 
 import io.micrometer.observation.ObservationRegistry;
 
@@ -13,11 +13,12 @@ import io.micrometer.observation.ObservationRegistry;
 @ComponentScan
 public class LayoutParsingServiceProcessorConfiguration {
 
+
     @Bean
     @Autowired
-    public PDFTronViewerDocumentService viewerDocumentService(ObservationRegistry registry) {
+    public ViewerDocumentService viewerDocumentService(ObservationRegistry registry) {
 
-        return new PDFTronViewerDocumentService(registry);
+        return new ViewerDocumentService(registry);
     }
 
 }

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

@@ -1,19 +1,15 @@
 package com.knecon.fforesight.service.layoutparser.processor;
 
-import java.io.ByteArrayInputStream;
 import java.io.File;
 import java.io.FileInputStream;
 import java.io.IOException;
 import java.io.InputStream;
-import java.nio.charset.StandardCharsets;
 import java.nio.file.Files;
 import java.nio.file.Path;
 import java.nio.file.Paths;
 import java.nio.file.StandardOpenOption;
 import java.util.Optional;
-import java.util.concurrent.CompletableFuture;
 
-import org.springframework.core.task.TaskExecutor;
 import org.springframework.stereotype.Service;
 
 import com.fasterxml.jackson.databind.ObjectMapper;
@@ -25,7 +21,6 @@ import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsi
 import com.knecon.fforesight.service.layoutparser.processor.python_api.model.image.ImageServiceResponse;
 import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableServiceResponse;
 import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.VisualLayoutParsingResponse;
-import com.knecon.fforesight.service.viewerdoc.service.ViewerDocVersioningUtility;
 import com.knecon.fforesight.tenantcommons.TenantContext;
 
 import io.micrometer.observation.annotation.Observed;
@@ -41,9 +36,6 @@ public class LayoutParsingStorageService {
     private final StorageService storageService;
     private final ObjectMapper objectMapper;
 
-    private final TaskExecutor taskExecutor;
-
-
     @Observed(name = "LayoutParsingStorageService", contextualName = "get-origin-file")
     public File getOriginFile(String storageId) throws IOException {
 
@@ -61,18 +53,11 @@ public class LayoutParsingStorageService {
         }
         File tempFile = createTempFile("viewerDocument", ".pdf");
         storageService.downloadTo(TenantContext.getTenantId(), storageId, tempFile);
-
-        if (!ViewerDocVersioningUtility.isCurrentVersion(tempFile)) {
-            assert tempFile.delete();
-            return Optional.empty();
-        }
-
         return Optional.of(tempFile);
     }
 
 
-    @SneakyThrows
-    public ImageServiceResponse getImagesFile(String storageId) {
+    public ImageServiceResponse getImagesFile(String storageId) throws IOException {
 
         try (InputStream inputStream = getObject(storageId)) {
 
@@ -83,8 +68,7 @@ public class LayoutParsingStorageService {
     }
 
 
-    @SneakyThrows
-    public TableServiceResponse getTablesFile(String storageId) {
+    public TableServiceResponse getTablesFile(String storageId) throws IOException {
 
         try (var tableClassificationStream = getObject(storageId)) {
 
@@ -94,45 +78,22 @@ public class LayoutParsingStorageService {
         }
     }
 
-
-@SneakyThrows
-    public VisualLayoutParsingResponse getVisualLayoutParsingFile(String storageId) {
+    public VisualLayoutParsingResponse getVisualLayoutParsingFile(String storageId) throws IOException {
 
         try (InputStream inputStream = getObject(storageId)) {
-            return objectMapper.readValue(inputStream, VisualLayoutParsingResponse.class);
+            VisualLayoutParsingResponse visualLayoutParsingResponse = objectMapper.readValue(inputStream, VisualLayoutParsingResponse.class);
+            return visualLayoutParsingResponse;
         }
     }
 
 
-    @SneakyThrows
     @Observed(name = "LayoutParsingStorageService", contextualName = "store-document-data")
     public void storeDocumentData(LayoutParsingRequest layoutParsingRequest, DocumentData documentData) {
 
-        Runnable storeDocumentStructureRunnable = () -> storageService.storeProtoObject(TenantContext.getTenantId(),
-                                                                                       layoutParsingRequest.structureFileStorageId(),
-                                                                                       documentData.getDocumentStructure());
-
-        CompletableFuture<Void> storeDocumentStructureFuture = CompletableFuture.runAsync(storeDocumentStructureRunnable, taskExecutor);
-
-        Runnable storeDocumentTextDataRunnable = () -> storageService.storeProtoObject(TenantContext.getTenantId(),
-                                                                                      layoutParsingRequest.textBlockFileStorageId(),
-                                                                                      documentData.getDocumentTextData());
-
-        CompletableFuture<Void> storeDocumentTextDataFuture = CompletableFuture.runAsync(storeDocumentTextDataRunnable, taskExecutor);
-
-        Runnable storeDocumentPositionsRunnable = () -> storageService.storeProtoObject(TenantContext.getTenantId(),
-                                                                                       layoutParsingRequest.positionBlockFileStorageId(),
-                                                                                       documentData.getDocumentPositions());
-
-        CompletableFuture<Void> storeDocumentPositionsFuture = CompletableFuture.runAsync(storeDocumentPositionsRunnable, taskExecutor);
-
-        Runnable storeDocumentPagesRunnable = () -> storageService.storeProtoObject(TenantContext.getTenantId(),
-                                                                                   layoutParsingRequest.pageFileStorageId(),
-                                                                                   documentData.getDocumentPages());
-
-        CompletableFuture<Void> storeDocumentPagesFuture = CompletableFuture.runAsync(storeDocumentPagesRunnable, taskExecutor);
-
-        CompletableFuture.allOf(storeDocumentStructureFuture, storeDocumentTextDataFuture, storeDocumentPositionsFuture, storeDocumentPagesFuture).join();
+        storageService.storeJSONObject(TenantContext.getTenantId(), layoutParsingRequest.structureFileStorageId(), documentData.getDocumentStructure());
+        storageService.storeJSONObject(TenantContext.getTenantId(), layoutParsingRequest.textBlockFileStorageId(), documentData.getDocumentTextData());
+        storageService.storeJSONObject(TenantContext.getTenantId(), layoutParsingRequest.positionBlockFileStorageId(), documentData.getDocumentPositions());
+        storageService.storeJSONObject(TenantContext.getTenantId(), layoutParsingRequest.pageFileStorageId(), documentData.getDocumentPages());
     }
 
 
@@ -141,6 +102,11 @@ public class LayoutParsingStorageService {
         storageService.storeJSONObject(TenantContext.getTenantId(), layoutParsingRequest.researchDocumentStorageId(), researchDocumentData);
     }
 
+    public void storeObject(String storageId, InputStream in) {
+
+        storageService.storeObject(TenantContext.getTenantId(), storageId, in);
+    }
+
 
     private File createTempFile(String filenamePrefix, String filenameSuffix) throws IOException {
 
@@ -193,16 +159,4 @@ public class LayoutParsingStorageService {
         }
     }
 
-
-    @SneakyThrows
-    @Observed(name = "LayoutParsingStorageService", contextualName = "store-markdown-file")
-    public void storeMarkdownFile(String markdownFileStorageId, String markdownContent) {
-
-        try (InputStream inputStream = new ByteArrayInputStream(markdownContent.getBytes(StandardCharsets.UTF_8))) {
-
-            storageService.storeObject(TenantContext.getTenantId(), markdownFileStorageId, inputStream);
-        }
-
-    }
-
 }

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

@@ -17,5 +17,4 @@ public class LayoutparserSettings {
 
     boolean debug;
     LayoutParsingType layoutParsingTypeOverride;
-    String pdftronLicense;
 }

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

@@ -0,0 +1,74 @@
+package com.knecon.fforesight.service.layoutparser.processor;
+
+import java.nio.charset.StandardCharsets;
+import java.util.List;
+
+import org.commonmark.Extension;
+import org.commonmark.ext.gfm.tables.TablesExtension;
+import org.commonmark.node.Document;
+import org.commonmark.node.Node;
+import org.commonmark.parser.Parser;
+import org.commonmark.renderer.Renderer;
+import org.commonmark.renderer.markdown.MarkdownRenderer;
+import org.springframework.stereotype.Service;
+
+import com.iqser.red.storage.commons.service.StorageService;
+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.markdown.MarkdownChunker;
+import com.knecon.fforesight.tenantcommons.TenantContext;
+
+import io.micrometer.observation.annotation.Observed;
+import lombok.AccessLevel;
+import lombok.RequiredArgsConstructor;
+import lombok.SneakyThrows;
+import lombok.experimental.FieldDefaults;
+
+@Service
+@RequiredArgsConstructor
+@FieldDefaults(makeFinal = true, level = AccessLevel.PRIVATE)
+public class MarkdownParsingPipeline {
+
+    StorageService storageService;
+
+
+    @SneakyThrows
+    @Observed(name = "MarkdownParsingPipeline", contextualName = "parse-markdown")
+    public LayoutParsingFinishedEvent parseMarkdownAndSaveToStorage(LayoutParsingRequest request) {
+
+        long start = System.currentTimeMillis();
+        String markdown;
+        try (var in = storageService.getObject(TenantContext.getTenantId(), request.originFileStorageId()).getInputStream()) {
+            markdown = new String(in.readAllBytes(), StandardCharsets.UTF_8);
+        }
+        Parser parser = buildParser();
+        Node node = parser.parse(markdown);
+
+        MarkdownChunker chunker = new MarkdownChunker(600);
+
+        node.accept(chunker);
+
+        Renderer renderer = buildRenderer();
+        List<Document> markdownChunks = chunker.getResult();
+        for (Document markdownChunk : markdownChunks) {
+
+        }
+
+        return LayoutParsingFinishedEvent.builder().identifier(request.identifier()).numberOfPages(1).duration(System.currentTimeMillis() - start).build();
+    }
+
+
+    public static Parser buildParser() {
+
+        List<Extension> extensions = List.of(TablesExtension.create());
+        return Parser.builder().extensions(extensions).build();
+    }
+
+
+    public static MarkdownRenderer buildRenderer() {
+
+        List<Extension> extensions = List.of(TablesExtension.create());
+        return MarkdownRenderer.builder().extensions(extensions).build();
+    }
+
+}

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

@@ -1,7 +1,6 @@
 package com.knecon.fforesight.service.layoutparser.processor.docstrum;
 
 import java.util.ArrayList;
-import java.util.EnumMap;
 import java.util.List;
 import java.util.stream.Collectors;
 
@@ -18,8 +17,8 @@ import com.knecon.fforesight.service.layoutparser.processor.docstrum.service.Zon
 import com.knecon.fforesight.service.layoutparser.processor.model.table.CleanRulings;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.RedTextPosition;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.TextDirection;
-import com.knecon.fforesight.service.layoutparser.processor.model.text.Word;
-import com.knecon.fforesight.service.layoutparser.processor.visualization.LayoutDebugLayer;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPositionSequence;
+import com.knecon.fforesight.service.layoutparser.processor.visualization.LayoutparsingVisualizations;
 
 import lombok.RequiredArgsConstructor;
 
@@ -27,7 +26,6 @@ import lombok.RequiredArgsConstructor;
 @RequiredArgsConstructor
 public class DocstrumSegmentationService {
 
-    public static final double SAME_DIRECTION_THRESHOLD = 0.9;
     private final NearestNeighbourService nearestNeighbourService;
     private final SpacingService spacingService;
     private final LineBuilderService lineBuilderService;
@@ -35,57 +33,30 @@ public class DocstrumSegmentationService {
     private final ReadingOrderService readingOrderService;
 
 
-    public List<Zone> segmentPage(List<Word> textPositions, boolean xyOrder, CleanRulings usedRulings, LayoutDebugLayer visualizations) {
+    public List<Zone> segmentPage(List<TextPositionSequence> textPositions, boolean xyOrder, CleanRulings usedRulings, LayoutparsingVisualizations visualizations) {
 
-        EnumMap<TextDirection, Integer> directionCounts = new EnumMap<>(TextDirection.class);
+        List<Zone> zones = new ArrayList<>();
+        zones.addAll(computeZones(textPositions, usedRulings, visualizations, TextDirection.ZERO));
+        zones.addAll(computeZones(textPositions, usedRulings, visualizations, TextDirection.QUARTER_CIRCLE));
+        zones.addAll(computeZones(textPositions, usedRulings, visualizations, TextDirection.HALF_CIRCLE));
+        zones.addAll(computeZones(textPositions, usedRulings, visualizations, TextDirection.THREE_QUARTER_CIRCLE));
 
-        List<Zone> newZones = computeZones(textPositions, usedRulings, visualizations, TextDirection.ZERO);
-        directionCounts.put(TextDirection.ZERO, newZones.size());
-        List<Zone> zones = new ArrayList<>(newZones);
-
-        newZones = computeZones(textPositions, usedRulings, visualizations, TextDirection.QUARTER_CIRCLE);
-        directionCounts.put(TextDirection.QUARTER_CIRCLE, newZones.size());
-        zones.addAll(newZones);
-
-        newZones = computeZones(textPositions, usedRulings, visualizations, TextDirection.HALF_CIRCLE);
-        directionCounts.put(TextDirection.HALF_CIRCLE, newZones.size());
-        zones.addAll(newZones);
-
-        newZones = computeZones(textPositions, usedRulings, visualizations, TextDirection.THREE_QUARTER_CIRCLE);
-        directionCounts.put(TextDirection.THREE_QUARTER_CIRCLE, newZones.size());
-        zones.addAll(newZones);
-
-        return readingOrderService.resolve(zones, xyOrder, mostSameDirection(directionCounts));
+        return readingOrderService.resolve(zones, xyOrder);
     }
 
 
-    private boolean mostSameDirection(EnumMap<TextDirection, Integer> directionCounts) {
+    private List<Zone> computeZones(List<TextPositionSequence> textPositions, CleanRulings rulings, LayoutparsingVisualizations visualizations, TextDirection direction) {
 
-        int total = directionCounts.values()
-                .stream()
-                .mapToInt(i -> i).sum();
-
-        if ((double) directionCounts.get(TextDirection.ZERO) / total > SAME_DIRECTION_THRESHOLD) {
-            return true;
-        } else if ((double) directionCounts.get(TextDirection.QUARTER_CIRCLE) / total > SAME_DIRECTION_THRESHOLD) {
-            return true;
-        } else if ((double) directionCounts.get(TextDirection.HALF_CIRCLE) / total > SAME_DIRECTION_THRESHOLD) {
-            return true;
-        } else if ((double) directionCounts.get(TextDirection.THREE_QUARTER_CIRCLE) / total > SAME_DIRECTION_THRESHOLD) {
-            return true;
-        }
-        return false;
-    }
-
-
-    private List<Zone> computeZones(List<Word> textPositions, CleanRulings rulings, LayoutDebugLayer visualizations, TextDirection direction) {
-
-        List<Character> characters = textPositions.stream()
+        List<RedTextPosition> positions = textPositions.stream()
                 .filter(t -> t.getDir() == direction)
-                .map(Word::getCharacters)
+                .map(TextPositionSequence::getTextPositions)
                 .flatMap(List::stream)
                 .toList();
 
+        List<Character> characters = positions.stream()
+                .map(Character::new)
+                .collect(Collectors.toList());
+
         nearestNeighbourService.findNearestNeighbors(characters);
 
         double characterSpacing = spacingService.computeCharacterSpacing(characters);

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/docstrum/model/AngleFilter.java

@@ -15,16 +15,10 @@ public class AngleFilter {
 
     public boolean matches(Neighbor neighbor) {
 
-        return matches(neighbor.getAngle());
-    }
-
-
-    public boolean matches(double angle) {
-
         if (lowerAngle <= upperAngle) {
-            return lowerAngle <= angle && angle < upperAngle;
+            return lowerAngle <= neighbor.getAngle() && neighbor.getAngle() < upperAngle;
         } else {
-            return lowerAngle <= angle || angle < upperAngle;
+            return lowerAngle <= neighbor.getAngle() || neighbor.getAngle() < upperAngle;
         }
     }
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/docstrum/model/BoundingBox.java

@@ -7,12 +7,8 @@ import java.util.List;
 import com.knecon.fforesight.service.layoutparser.processor.utils.RectangleTransformations;
 
 import lombok.Data;
-import lombok.NoArgsConstructor;
-import lombok.experimental.SuperBuilder;
 
 @Data
-@SuperBuilder
-@NoArgsConstructor
 public abstract class BoundingBox {
 
     // Java coordinate system: (0, 0) is always upper left, x is increasing left to right and y is increasing from top to bottom.
@@ -23,7 +19,7 @@ public abstract class BoundingBox {
     // This rotates completely in 90 degree steps with page rotation.
     // Needs to be used when writing to a PDF.
     // Also, these are definitely correct and should be used whenever possible.
-    protected Rectangle2D bBoxPdf;
+    protected Rectangle2D bBoxInitialUserSpace;
 
     protected static final float VERTICAL_COMPARISON_THRESHOLD = 0.4f;
 
@@ -54,25 +50,25 @@ public abstract class BoundingBox {
 
     public double getPdfMinX() {
 
-        return bBoxPdf.getMinX();
+        return bBoxInitialUserSpace.getMinX();
     }
 
 
     public double getPdfMaxX() {
 
-        return bBoxPdf.getMaxX();
+        return bBoxInitialUserSpace.getMaxX();
     }
 
 
     public double getPdfMinY() {
 
-        return bBoxPdf.getMinY();
+        return bBoxInitialUserSpace.getMinY();
     }
 
 
     public double getPdfMaxY() {
 
-        return bBoxPdf.getMaxY();
+        return bBoxInitialUserSpace.getMaxY();
     }
 
 
@@ -133,31 +129,13 @@ public abstract class BoundingBox {
     }
 
 
-    public boolean intersectsX(BoundingBox other, float threshold) {
-
-        return this.getX() - threshold <= other.getMaxX() && this.getMaxX() + threshold >= other.getX();
-    }
-
-
-    public boolean intersectsPdf(BoundingBox other) {
-
-        return this.intersectsXPdf(other) && this.intersectsYPdf(other);
-    }
-
-
-    public boolean intersectsPdf(BoundingBox other, float yThreshold, float xThreshold) {
-
-        return this.intersectsXPdf(other, xThreshold) && this.intersectsYPdf(other, yThreshold);
-    }
-
-
-    public boolean intersectsYPdf(BoundingBox other) {
+    public boolean intersectsY(BoundingBox other) {
 
         return this.getPdfMinY() <= other.getPdfMaxY() && this.getPdfMaxY() >= other.getPdfMinY();
     }
 
 
-    public boolean intersectsY(BoundingBox other) {
+    public boolean intersectsYJava(BoundingBox other) {
 
         return this.getY() <= other.getMaxY() && this.getMaxY() >= other.getY();
     }
@@ -165,31 +143,25 @@ public abstract class BoundingBox {
 
     public boolean intersectsY(BoundingBox other, float threshold) {
 
-        return this.getY() - threshold <= other.getMaxY() && this.getMaxY() + threshold >= other.getY();
-    }
-
-
-    public boolean intersectsYPdf(BoundingBox other, float threshold) {
-
         return this.getPdfMinY() - threshold <= other.getPdfMaxY() && this.getPdfMaxY() + threshold >= other.getPdfMinY();
     }
 
 
-    public boolean intersectsXPdf(BoundingBox other) {
-
-        return this.getPdfMinX() <= other.getPdfMaxX() && this.getPdfMaxX() >= other.getPdfMinX();
-    }
-
-
     public boolean intersectsX(BoundingBox other) {
 
+        return this.getPdfMinX() <= other.getPdfMaxX() && this.getPdfMaxX() >= other.getPdfMinX();
+    }
+
+
+    public boolean intersectsXJava(BoundingBox other) {
+
         return this.getX() <= other.getMaxX() && this.getMaxX() >= other.getMinX();
     }
 
 
-    public boolean intersectsXPdf(BoundingBox other, float threshold) {
+    public boolean intersectsX(BoundingBox other, float threshold) {
 
-        return this.getPdfMinX() - threshold <= other.getPdfMaxX() && this.getMaxX() + threshold >= other.getPdfMinX();
+        return this.getPdfMinX() - threshold <= other.getMaxX() && this.getMaxX() + threshold >= other.getPdfMinX();
     }
 
 
@@ -198,8 +170,8 @@ public abstract class BoundingBox {
         this.bBox = components.stream()
                 .map(BoundingBox::getBBox)
                 .collect(RectangleTransformations.collectBBox());
-        this.bBoxPdf = components.stream()
-                .map(BoundingBox::getBBoxPdf)
+        this.bBoxInitialUserSpace = components.stream()
+                .map(BoundingBox::getBBoxInitialUserSpace)
                 .collect(RectangleTransformations.collectBBox());
     }
 
@@ -225,55 +197,57 @@ public abstract class BoundingBox {
 
     public double horizontalDistance(BoundingBox other) {
 
-        double rect1Right = getMaxX();
-        double rect1Left = getMinX();
-        double rect2Right = other.getMaxX();
-        double rect2Left = other.getMinX();
-
-        if (rect1Left > rect2Right || rect2Left > rect1Right) {
-            return Math.max(rect2Left - rect1Right, rect1Left - rect2Right);
+        Rectangle2D left;
+        Rectangle2D right;
+        if (this.leftOf(other)) {
+            left = this.getBBox();
+            right = other.getBBox();
         } else {
-            return 0;
+            left = other.getBBox();
+            right = this.getBBox();
         }
+
+        return Math.max(0, right.getMinX() - left.getMaxX());
     }
 
 
     public double verticalDistance(BoundingBox other) {
 
-        double rect1Top = getMaxY();
-        double rect1Bottom = getMinY();
-        double rect2Top = other.getMaxY();
-        double rect2Bottom = other.getMinY();
-
-        if (rect1Bottom > rect2Top || rect2Bottom > rect1Top) {
-            return Math.max(rect2Bottom - rect1Top, rect1Bottom - rect2Top);
+        Rectangle2D bottom;
+        Rectangle2D top;
+        if (this.isAbove(other)) {
+            top = this.getBBox();
+            bottom = other.getBBox();
         } else {
-            return 0;
+            bottom = this.getBBox();
+            top = other.getBBox();
         }
+
+        return Math.max(0, bottom.getMinY() - top.getMaxY());
     }
 
 
     public boolean rightOf(BoundingBox other) {
 
-        return this.intersectsY(other) && other.getMaxX() <= this.getMinX();
+        return this.intersectsYJava(other) && other.getMaxX() <= this.getMinX();
     }
 
 
     public boolean leftOf(BoundingBox other) {
 
-        return this.intersectsY(other) && other.getMinX() >= this.getMaxX();
+        return this.intersectsYJava(other) && other.getMinX() >= this.getMaxX();
     }
 
 
     public boolean isAbove(BoundingBox other) {
 
-        return this.intersectsX(other) && other.getMinY() >= this.getMaxY();
+        return this.intersectsXJava(other) && other.getMinY() >= this.getMaxY();
     }
 
 
     public boolean isBelow(BoundingBox other) {
 
-        return this.intersectsX(other) && this.getMinY() >= other.getMaxY();
+        return this.intersectsXJava(other) && this.getMinY() >= other.getMaxY();
     }
 
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/docstrum/model/Character.java

@@ -35,7 +35,7 @@ public class Character {
 
     public double getHeight() {
 
-        return textPosition.getHeightDirAdj();
+        return textPosition.getHeightDir();
     }
 
 
@@ -65,9 +65,9 @@ public class Character {
         double s = Math.sin(-0);
         double c = Math.cos(-0);
         xs[0] = c * x - s * y;
-        xs[1] = c * (x + textPosition.getWidthDirAdj()) - s * (y + textPosition.getHeightDirAdj());
+        xs[1] = c * (x + textPosition.getWidthDirAdj()) - s * (y + textPosition.getHeightDir());
         xs[2] = c * other.x - s * other.y;
-        xs[3] = c * (other.x + other.textPosition.getWidthDirAdj()) - s * (other.y + other.textPosition.getHeightDirAdj());
+        xs[3] = c * (other.x + other.textPosition.getWidthDirAdj()) - s * (other.y + other.textPosition.getHeightDir());
         boolean overlapping = xs[1] >= xs[2] && xs[3] >= xs[0];
         Arrays.sort(xs);
         return Math.abs(xs[2] - xs[1]) * (overlapping ? 1 : -1);

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/docstrum/model/Line.java

@@ -1,30 +1,20 @@
 package com.knecon.fforesight.service.layoutparser.processor.docstrum.model;
 
-import static com.knecon.fforesight.service.layoutparser.processor.model.text.Word.BOLD;
-import static com.knecon.fforesight.service.layoutparser.processor.model.text.Word.BOLD_ITALIC;
-import static com.knecon.fforesight.service.layoutparser.processor.model.text.Word.ITALIC;
-import static com.knecon.fforesight.service.layoutparser.processor.model.text.Word.STANDARD;
-
 import java.util.ArrayList;
 import java.util.Arrays;
-import java.util.Comparator;
-import java.util.EnumMap;
 import java.util.List;
-import java.util.Map;
-import java.util.concurrent.atomic.AtomicInteger;
 
 import com.knecon.fforesight.service.layoutparser.processor.docstrum.utils.FastAtan2;
-import com.knecon.fforesight.service.layoutparser.processor.model.text.Word;
-import com.knecon.fforesight.service.layoutparser.processor.model.text.FontStyle;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPositionSequence;
 
 import lombok.Data;
 import lombok.EqualsAndHashCode;
 
 @Data
-@EqualsAndHashCode(onlyExplicitlyIncluded = true, callSuper = false)
-public class Line extends TextBoundingBox {
+@EqualsAndHashCode(onlyExplicitlyIncluded = true)
+public class Line extends BoundingBox {
 
-    private static final double WORD_DISTANCE_MULTIPLIER = 0.17;
+    private static final double WORD_DISTANCE_MULTIPLIER = 0.18;
 
     @EqualsAndHashCode.Include
     private final double x0;
@@ -36,13 +26,16 @@ public class Line extends TextBoundingBox {
     @EqualsAndHashCode.Include
     private final double y1;
 
-    private FontStyle fontStyle;
+    private final double height;
 
-    private final List<Word> words;
+    private final List<Character> characters;
+    private final List<TextPositionSequence> words = new ArrayList<>();
 
 
     public Line(List<Character> characters, double wordSpacing) {
 
+        this.characters = characters;
+
         if (characters.size() >= 2) {
             // linear regression
             double sx = 0.0;
@@ -71,43 +64,9 @@ public class Line extends TextBoundingBox {
             this.y0 = character.getY() - dy;
             this.y1 = character.getY() + dy;
         }
-        this.words = new ArrayList<>();
-        computeWords(characters, wordSpacing * WORD_DISTANCE_MULTIPLIER);
+        height = computeHeight();
+        computeWords(wordSpacing * WORD_DISTANCE_MULTIPLIER);
         buildBBox();
-        computeFontStyle();
-    }
-
-
-    public Line(List<Word> words) {
-
-        this.words = words;
-        buildBBox();
-        x0 = getMinX();
-        y0 = getMinY();
-        x1 = getMaxX();
-        y1 = getMaxY();
-        computeFontStyle();
-    }
-
-
-    private void computeFontStyle() {
-
-        EnumMap<FontStyle, AtomicInteger> fontStyleCounter = new EnumMap<>(FontStyle.class);
-        for (FontStyle fontStyle : FontStyle.values()) {
-            fontStyleCounter.put(fontStyle, new AtomicInteger(0));
-        }
-        for (Word word : words) {
-            switch (word.getFontStyle()) {
-                case STANDARD -> fontStyleCounter.get(FontStyle.REGULAR).getAndIncrement();
-                case BOLD -> fontStyleCounter.get(FontStyle.BOLD).getAndIncrement();
-                case ITALIC -> fontStyleCounter.get(FontStyle.ITALIC).getAndIncrement();
-                case BOLD_ITALIC -> fontStyleCounter.get(FontStyle.BOLD_ITALIC).getAndIncrement();
-            }
-        }
-        fontStyle = fontStyleCounter.entrySet()
-                .stream()
-                .max(Comparator.comparing(entry -> entry.getValue().get()))
-                .map(Map.Entry::getKey).orElse(FontStyle.REGULAR);
     }
 
 
@@ -123,6 +82,14 @@ public class Line extends TextBoundingBox {
     }
 
 
+    private double computeHeight() {
+
+        return characters.stream()
+                       .map(Character::getHeight)
+                       .reduce(0d, Double::sum) / characters.size();
+    }
+
+
     public double angularDifference(Line j) {
 
         double diff = Math.abs(getAngle() - j.getAngle());
@@ -155,22 +122,19 @@ public class Line extends TextBoundingBox {
     }
 
 
-    private void computeWords(List<Character> characters, double wordSpacing) {
+    private void computeWords(double wordSpacing) {
 
-        // Imo, the width of space should be scaled with the font size, but it only depends on the median distance between horizontal neighbours.
-        // If there are large differences in fontsize on a page, this might lead to missing spaces for the smaller fonts and too many for larger fonts.
-        // I've just now changed the scaling factor. If you come across this comment with missing whitespaces again, try scaling the fontsize instead of simply changing the factor again.
-        Word word = new Word();
+        TextPositionSequence word = new TextPositionSequence();
         Character previous = null;
         for (Character current : characters) {
             if (previous != null) {
                 double dist = current.getTextPosition().getXDirAdj() - previous.getTextPosition().getXDirAdj() - previous.getTextPosition().getWidthDirAdj();
                 if (dist > wordSpacing) {
                     words.add(word);
-                    word = new Word();
+                    word = new TextPositionSequence();
                 }
             }
-            word.add(current);
+            word.getTextPositions().add(current.getTextPosition());
             previous = current;
         }
         words.add(word);
@@ -179,7 +143,9 @@ public class Line extends TextBoundingBox {
 
     private void buildBBox() {
 
-        this.setToBBoxOfComponents(words);
+        this.setToBBoxOfComponents(characters.stream()
+                                       .map(Character::getTextPosition)
+                                       .toList());
     }
 
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/docstrum/model/TextBoundingBox.java

@@ -1,180 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.processor.docstrum.model;
-
-import java.awt.geom.Rectangle2D;
-import java.util.List;
-import java.util.Set;
-import java.util.stream.Collectors;
-
-import com.knecon.fforesight.service.layoutparser.processor.model.text.TextDirection;
-import com.knecon.fforesight.service.layoutparser.processor.utils.RectangleTransformations;
-
-import lombok.EqualsAndHashCode;
-import lombok.Getter;
-import lombok.NoArgsConstructor;
-import lombok.Setter;
-import lombok.experimental.SuperBuilder;
-
-@Getter
-@Setter
-@SuperBuilder
-@NoArgsConstructor
-@EqualsAndHashCode(callSuper = false)
-public abstract class TextBoundingBox extends BoundingBox {
-
-    protected Rectangle2D bBoxDirAdj;
-
-    protected TextDirection dir;
-
-
-    @Override
-    public void setToBBoxOfComponents(List<? extends BoundingBox> components) {
-
-        super.setToBBoxOfComponents(components);
-        this.bBoxDirAdj = components.stream()
-                .filter(c -> c instanceof TextBoundingBox)
-                .map(c -> (TextBoundingBox) c)
-                .map(TextBoundingBox::getBBoxDirAdj)
-                .collect(RectangleTransformations.collectBBox());
-
-        Set<TextDirection> textDirections = components.stream()
-                .filter(c -> c instanceof TextBoundingBox)
-                .map(c -> (TextBoundingBox) c)
-                .map(TextBoundingBox::getDir)
-                .collect(Collectors.toSet());
-
-        if (textDirections.isEmpty()) {
-            dir = TextDirection.ZERO;
-        } else if (textDirections.size() > 1) {
-            throw new IllegalArgumentException("More than one text direction found");
-        } else {
-            dir = textDirections.iterator().next();
-        }
-    }
-
-
-    public double getXDirAdj() {
-
-        return this.bBoxDirAdj.getX();
-    }
-
-
-    public double getYDirAdj() {
-
-        return this.bBoxDirAdj.getY();
-    }
-
-
-    public double getWidthDirAdj() {
-
-        return this.bBoxDirAdj.getWidth();
-    }
-
-
-    public double getHeightDirAdj() {
-
-        return this.bBoxDirAdj.getHeight();
-    }
-
-
-    public double getMaxXDirAdj() {
-
-        return this.bBoxDirAdj.getMaxX();
-    }
-
-
-    public double getMaxYDirAdj() {
-
-        return this.bBoxDirAdj.getMaxY();
-    }
-
-
-    public double getCenterYDirAdj() {
-
-        return this.bBoxDirAdj.getCenterY();
-    }
-
-
-    public double getCenterXDirAdj() {
-
-        return this.bBoxDirAdj.getCenterX();
-    }
-
-
-    public double horizontalDistanceDirAdj(TextBoundingBox other) {
-
-        double rect1Right = getMaxXDirAdj();
-        double rect1Left = getXDirAdj();
-        double rect2Right = other.getMaxXDirAdj();
-        double rect2Left = other.getXDirAdj();
-
-        if (rect1Left > rect2Right || rect2Left > rect1Right) {
-            return Math.max(rect2Left - rect1Right, rect1Left - rect2Right);
-        } else {
-            return 0;
-        }
-    }
-
-
-    public double verticalDistanceDirAdj(TextBoundingBox other) {
-
-        double rect1Top = getMaxYDirAdj();
-        double rect1Bottom = getYDirAdj();
-        double rect2Top = other.getMaxYDirAdj();
-        double rect2Bottom = other.getYDirAdj();
-
-        if (rect1Bottom > rect2Top || rect2Bottom > rect1Top) {
-            return Math.max(rect2Bottom - rect1Top, rect1Bottom - rect2Top);
-        } else {
-            return 0;
-        }
-    }
-
-
-    public boolean intersectsDirAdj(TextBoundingBox other) {
-
-        return this.intersectsXDirAdj(other) && this.intersectsYDirAdj(other);
-    }
-
-
-    public boolean intersectsDirAdj(TextBoundingBox other, float yThreshold, float xThreshold) {
-
-        return this.intersectsXDirAdj(other, xThreshold) && this.intersectsYDirAdj(other, yThreshold);
-    }
-
-
-    public boolean intersectsXDirAdj(TextBoundingBox other, float threshold) {
-
-        return this.getXDirAdj() - threshold <= other.getMaxXDirAdj() && this.getMaxXDirAdj() + threshold >= other.getXDirAdj();
-    }
-
-
-    public boolean intersectsXDirAdj(TextBoundingBox other) {
-
-        return this.getXDirAdj() <= other.getMaxXDirAdj() && this.getMaxXDirAdj() >= other.getXDirAdj();
-    }
-
-
-    public boolean intersectsYDirAdj(TextBoundingBox other) {
-
-        return this.getYDirAdj() <= other.getMaxYDirAdj() && this.getMaxYDirAdj() >= other.getYDirAdj();
-    }
-
-
-    public boolean intersectsYDirAdj(TextBoundingBox other, float threshold) {
-
-        return this.getYDirAdj() - threshold <= other.getMaxYDirAdj() && this.getMaxYDirAdj() + threshold >= other.getYDirAdj();
-    }
-
-
-    public boolean isAboveDirAdj(TextBoundingBox other) {
-
-        return other.isBelow(this);
-    }
-
-
-    public boolean isBelowDirAdj(TextBoundingBox other) {
-
-        return this.intersectsXDirAdj(other) && this.getYDirAdj() >= other.getMaxYDirAdj();
-    }
-
-}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/docstrum/model/UnionFind.java

@@ -28,10 +28,4 @@ public class UnionFind<T> extends org.jgrapht.alg.util.UnionFind<T> {
         return setRep.values();
     }
 
-
-    public Collection<T> getElements() {
-
-        return getParentMap().keySet();
-    }
-
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/docstrum/model/Zone.java

@@ -6,11 +6,9 @@ import java.util.List;
 import com.knecon.fforesight.service.layoutparser.processor.utils.RectangleTransformations;
 
 import lombok.Data;
-import lombok.EqualsAndHashCode;
 
 @Data
-@EqualsAndHashCode(callSuper = false)
-public class Zone extends TextBoundingBox {
+public class Zone extends BoundingBox {
 
     private List<Line> lines;
 
@@ -18,6 +16,7 @@ public class Zone extends TextBoundingBox {
     @SuppressWarnings("PMD.ConstructorCallsOverridableMethod")
     public Zone(List<Line> lines) {
 
+        lines.sort(Comparator.comparingDouble(Line::getY0));
         this.lines = lines;
         setToBBoxOfComponents(lines);
     }

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

@@ -17,7 +17,7 @@ public class LineBuilderService {
 
     private static final double CHARACTER_SPACING_DISTANCE_MULTIPLIER = 3.5;
     private static final double LINE_SPACING_THRESHOLD_MULTIPLIER = 0.67;
-    private static final double ANGLE_TOLERANCE = Math.toRadians(5);
+    private static final double ANGLE_TOLERANCE = Math.PI / 6;
 
 
     public List<Line> buildLines(List<Character> characters, double characterSpacing, double lineSpacing, CleanRulings rulings) {

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

@@ -1,17 +1,15 @@
 package com.knecon.fforesight.service.layoutparser.processor.docstrum.service;
 
-import java.awt.geom.Rectangle2D;
 import java.util.ArrayList;
 import java.util.Comparator;
 import java.util.HashMap;
 import java.util.List;
 import java.util.ListIterator;
 import java.util.Map;
-import java.util.stream.Collectors;
 
 import org.springframework.stereotype.Service;
 
-import com.knecon.fforesight.service.layoutparser.processor.docstrum.model.TextBoundingBox;
+import com.knecon.fforesight.service.layoutparser.processor.docstrum.model.BoundingBox;
 import com.knecon.fforesight.service.layoutparser.processor.docstrum.model.Zone;
 import com.knecon.fforesight.service.layoutparser.processor.docstrum.utils.DoubleUtils;
 
@@ -21,30 +19,21 @@ public class ReadingOrderService {
     private static final double THRESHOLD = 5;
     public static final double MULTI_COLUMN_DETECTION_THRESHOLD = 1.5;
 
-    private static final Comparator<TextBoundingBox> COMPARATOR = //
-            Comparator.comparing(TextBoundingBox::getY, (o1, o2) -> DoubleUtils.compareDouble(o1, o2, THRESHOLD))
-                    .thenComparing(TextBoundingBox::getX, (o1, o2) -> DoubleUtils.compareDouble(o1, o2, THRESHOLD));
 
-    private static final Comparator<TextBoundingBox> COMPARATOR_DIR_ADJ = //
-            Comparator.comparing(TextBoundingBox::getYDirAdj, (o1, o2) -> DoubleUtils.compareDouble(o1, o2, THRESHOLD))
-                    .thenComparing(TextBoundingBox::getXDirAdj, (o1, o2) -> DoubleUtils.compareDouble(o1, o2, THRESHOLD));
-
-
-    public List<Zone> resolve(List<Zone> zones, boolean xyReadingOrder, boolean useDirAdjCoords) {
+    public List<Zone> resolve(List<Zone> zones, boolean xyReadingOrder) {
 
         if (zones.isEmpty() || zones.size() == 1) {
             return zones;
         }
 
         if (xyReadingOrder) {
-            return resolveSingleColumnReadingOrder(zones, useDirAdjCoords);
+            return resolveSingleColumnReadingOrder(zones);
         }
 
         Map<Long, Integer> histogram = new HashMap<>();
         for (Zone zone : zones) {
-            Rectangle2D bbox = useDirAdjCoords ? zone.getBBoxDirAdj() : zone.getBBox();
-            long minY = Math.round(bbox.getMinY());
-            long maxY = Math.round(bbox.getMaxY());
+            long minY = Math.round(zone.getBBox().getMinY());
+            long maxY = Math.round(zone.getBBox().getMaxY());
             for (long i = minY; i <= maxY; i++) {
                 histogram.put(i, histogram.getOrDefault(i, 0) + 1);
             }
@@ -54,32 +43,24 @@ public class ReadingOrderService {
                     .stream()
                     .mapToInt(Integer::intValue).average()
                     .orElse(1) < MULTI_COLUMN_DETECTION_THRESHOLD) {
-            return resolveSingleColumnReadingOrder(zones, useDirAdjCoords);
+            return resolveSingleColumnReadingOrder(zones);
         } else {
 
-            return resolveMultiColumnReadingOder(zones, useDirAdjCoords);
+            return resolveMultiColumnReadingOder(zones);
         }
 
     }
 
 
-    private static List<Zone> resolveSingleColumnReadingOrder(List<Zone> zones, boolean useDirAdjCoords) {
+    private static List<Zone> resolveSingleColumnReadingOrder(List<Zone> zones) {
 
-        if (useDirAdjCoords) {
-            return zones.stream()
-                    .collect(Collectors.groupingBy(TextBoundingBox::getDir)).values()
-                    .stream()
-                    .flatMap(words -> words.stream()
-                            .sorted(COMPARATOR_DIR_ADJ))
-                    .toList();
-        }
-
-        zones.sort(COMPARATOR);
+        zones.sort(Comparator.comparing(BoundingBox::getY, (o1, o2) -> DoubleUtils.compareDouble(o1, o2, THRESHOLD))
+                           .thenComparing(BoundingBox::getX, (o1, o2) -> DoubleUtils.compareDouble(o1, o2, THRESHOLD)));
         return zones;
     }
 
 
-    private List<Zone> resolveMultiColumnReadingOder(List<Zone> zones, boolean useDirAdjCoords) {
+    private List<Zone> resolveMultiColumnReadingOder(List<Zone> zones) {
 
         // Simple reading order resolver for multi column page layout as described here : https://pub.towardsai.net/advanced-rag-02-unveiling-pdf-parsing-b84ae866344e
         // TODO implement a more fancy reading order resolver see https://github.com/BobLd/DocumentLayoutAnalysis/blob/master/README.md#reading-order
@@ -88,12 +69,11 @@ public class ReadingOrderService {
         double maxX = Double.NEGATIVE_INFINITY;
 
         for (Zone zone : zones) {
-            Rectangle2D bbox = useDirAdjCoords ? zone.getBBoxDirAdj() : zone.getBBox();
-            if (bbox.getX() < minX) {
-                minX = zone.getXDirAdj();
+            if (zone.getX() < minX) {
+                minX = zone.getX();
             }
-            if (bbox.getMaxX() > maxX) {
-                maxX = zone.getMaxXDirAdj();
+            if (zone.getX() + zone.getWidth() > maxX) {
+                maxX = zone.getX() + zone.getWidth();
             }
         }
 
@@ -102,27 +82,24 @@ public class ReadingOrderService {
         List<Zone> leftOf = new ArrayList<>();
         List<Zone> rightOf = new ArrayList<>();
         List<Zone> middle = new ArrayList<>();
-
         for (Zone zone : zones) {
-            Rectangle2D bbox = useDirAdjCoords ? zone.getBBoxDirAdj() : zone.getBBox();
-            if (bbox.getX() < midLineXCoordinate && bbox.getX() + bbox.getWidth() < midLineXCoordinate) {
+            if (zone.getX() < midLineXCoordinate && zone.getX() + zone.getWidth() < midLineXCoordinate) {
                 leftOf.add(zone);
-            } else if (bbox.getX() > midLineXCoordinate && bbox.getX() + bbox.getWidth() > midLineXCoordinate) {
+            } else if (zone.getX() > midLineXCoordinate && zone.getX() + zone.getWidth() > midLineXCoordinate) {
                 rightOf.add(zone);
             } else {
                 middle.add(zone);
             }
         }
 
-        if (useDirAdjCoords) {
-            leftOf.sort(COMPARATOR_DIR_ADJ);
-            rightOf.sort(COMPARATOR_DIR_ADJ);
-            middle.sort(COMPARATOR_DIR_ADJ);
-        } else {
-            leftOf.sort(COMPARATOR);
-            rightOf.sort(COMPARATOR);
-            middle.sort(COMPARATOR);
-        }
+        leftOf.sort(Comparator.comparing(BoundingBox::getY, (o1, o2) -> DoubleUtils.compareDouble(o1, o2, THRESHOLD))
+                            .thenComparing(BoundingBox::getX, (o1, o2) -> DoubleUtils.compareDouble(o1, o2, THRESHOLD)));
+
+        rightOf.sort(Comparator.comparing(BoundingBox::getY, (o1, o2) -> DoubleUtils.compareDouble(o1, o2, THRESHOLD))
+                             .thenComparing(BoundingBox::getX, (o1, o2) -> DoubleUtils.compareDouble(o1, o2, THRESHOLD)));
+
+        middle.sort(Comparator.comparing(BoundingBox::getY, (o1, o2) -> DoubleUtils.compareDouble(o1, o2, THRESHOLD))
+                            .thenComparing(BoundingBox::getX, (o1, o2) -> DoubleUtils.compareDouble(o1, o2, THRESHOLD)));
 /*
         List<Zone> leftNotIntersecting = new ArrayList<>();
         for (Zone leftZone : leftOf) {
@@ -174,9 +151,8 @@ public class ReadingOrderService {
 
         while (itty.hasNext()) {
             Zone current = itty.next();
-            Rectangle2D bbox = useDirAdjCoords ? current.getBBoxDirAdj() : current.getBBox();
             for (int i = 0; i < sortedZones.size(); i++) {
-                if (bbox.getY() < sortedZones.get(i).getY()) {
+                if (current.getY() < sortedZones.get(i).getY()) {
                     sortedZones.add(i, current);
                     itty.remove();
                     break;

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

@@ -1,7 +1,7 @@
 package com.knecon.fforesight.service.layoutparser.processor.docstrum.service;
 
 import java.util.ArrayList;
-import java.util.Collection;
+import java.util.Comparator;
 import java.util.HashSet;
 import java.util.List;
 import java.util.Set;
@@ -9,12 +9,11 @@ import java.util.stream.Collectors;
 
 import org.springframework.stereotype.Service;
 
+import com.knecon.fforesight.service.layoutparser.processor.docstrum.model.Character;
 import com.knecon.fforesight.service.layoutparser.processor.docstrum.model.Line;
 import com.knecon.fforesight.service.layoutparser.processor.docstrum.model.UnionFind;
 import com.knecon.fforesight.service.layoutparser.processor.docstrum.model.Zone;
 import com.knecon.fforesight.service.layoutparser.processor.model.table.CleanRulings;
-import com.knecon.fforesight.service.layoutparser.processor.model.text.Word;
-import com.knecon.fforesight.service.layoutparser.processor.utils.TextPositionOperations;
 
 @Service
 public class ZoneBuilderService {
@@ -22,7 +21,7 @@ public class ZoneBuilderService {
     private static final double MIN_HORIZONTAL_DISTANCE_MULTIPLIER = -0.5;
     private static final double MAX_VERTICAL_DISTANCE_MULTIPLIER = 1.2;
 
-    private static final double MIN_HORIZONTAL_MERGE_DISTANCE_MULTIPLIER = -7;
+    private static final double MIN_HORIZONTAL_MERGE_DISTANCE_MULTIPLIER = -3.0;
 
     private static final double MAX_VERTICAL_MERGE_DISTANCE_MULTIPLIER = 0.5;
 
@@ -30,7 +29,7 @@ public class ZoneBuilderService {
 
     private static final double MAX_LINE_SIZE_SCALE = 2.5;
 
-    private static final double ANGLE_TOLERANCE = Math.toRadians(5);
+    private static final double ANGLE_TOLERANCE = Math.PI / 6;
 
     private static final double MAX_VERTICAL_MERGE_DISTANCE = 0.5;
 
@@ -39,7 +38,7 @@ public class ZoneBuilderService {
 
         double minHorizontalDistance = characterSpacing * MIN_HORIZONTAL_DISTANCE_MULTIPLIER;
         double maxVerticalDistance = lineSpacing * MAX_VERTICAL_DISTANCE_MULTIPLIER;
-        double minHorizontalMergeDistance = lineSpacing * MIN_HORIZONTAL_MERGE_DISTANCE_MULTIPLIER;
+        double minHorizontalMergeDistance = characterSpacing * MIN_HORIZONTAL_MERGE_DISTANCE_MULTIPLIER;
         double maxVerticalMergeDistance = lineSpacing * MAX_VERTICAL_MERGE_DISTANCE_MULTIPLIER;
 
         UnionFind<Line> unionFind = new UnionFind<>(new HashSet<>(lines));
@@ -55,26 +54,11 @@ public class ZoneBuilderService {
                     return;
                 }
 
-//                if (!innerLine.getFontStyle().equals(outerLine.getFontStyle()) //
-//                    && !outerLine.intersectsY(innerLine, -2f)) {
-//                    return;
-//                }
+                double scale = Math.min(outerLine.getHeight(), innerLine.getHeight()) / meanHeight;
+                scale = Math.max(MIN_LINE_SIZE_SCALE, Math.min(scale, MAX_LINE_SIZE_SCALE));
 
-                double horizontalScale = Math.min(outerLine.getHeightDirAdj(), innerLine.getHeightDirAdj()) / meanHeight;
-                horizontalScale = Math.max(MIN_LINE_SIZE_SCALE, Math.min(horizontalScale, MAX_LINE_SIZE_SCALE));
-                double verticalScale = horizontalScale;
-
-//                if (innerLine.toString().endsWith(":")
-//                    || outerLine.toString().endsWith(":")
-//                    || numericalIdentifierPattern.matcher(innerLine.toString()).matches()
-//                    || numericalIdentifierPattern.matcher(outerLine.toString()).matches()) {
-//
-//                    horizontalScale *= 5;
-//                    verticalScale /= 10;
-//                }
-
-                double horizontalDistance = outerLine.horizontalDistance(innerLine) / horizontalScale;
-                double verticalDistance = outerLine.verticalDistance(innerLine) / verticalScale;
+                double horizontalDistance = outerLine.horizontalDistance(innerLine) / scale;
+                double verticalDistance = outerLine.verticalDistance(innerLine) / scale;
 
                 if ((!(minHorizontalDistance <= horizontalDistance) || !(verticalDistance <= maxVerticalDistance)) //
                     && (!(minHorizontalMergeDistance <= horizontalDistance) || !(verticalDistance <= maxVerticalMergeDistance))) {
@@ -103,7 +87,7 @@ public class ZoneBuilderService {
         double weights = 0.0;
         for (Line line : lines) {
             double weight = line.getLength();
-            meanHeight += line.getHeightDirAdj() * weight;
+            meanHeight += line.getHeight() * weight;
             weights += weight;
         }
         meanHeight /= weights;
@@ -113,14 +97,64 @@ public class ZoneBuilderService {
 
     private Zone mergeLinesInZone(List<Line> lines, double characterSpacing, double lineSpacing) {
 
-        Set<Word> words = lines.stream()
-                .map(Line::getWords)
-                .flatMap(Collection::stream)
-                .collect(Collectors.toSet());
-        Collection<Set<Word>> groupedLines = TextPositionOperations.groupByLine(words);
+        double maxHorizontalDistance = 0;
+        double minVerticalDistance = 0;
+        double maxVerticalDistance = lineSpacing * MAX_VERTICAL_MERGE_DISTANCE;
 
-        List<Line> sortedLines = TextPositionOperations.sortLines(groupedLines);
-        return new Zone(sortedLines);
+        UnionFind<Line> unionFind = new UnionFind<>(new HashSet<>(lines));
+
+        lines.forEach(outer -> {
+            lines.forEach(inner -> {
+                if (inner == outer) {
+                    return;
+                }
+
+                double horizontalDistance = outer.horizontalDistance(inner);
+                double verticalDistance = outer.verticalDistance(inner);
+
+                if (horizontalDistance <= maxHorizontalDistance && minVerticalDistance <= verticalDistance && verticalDistance <= maxVerticalDistance) {
+
+                    unionFind.union(outer, inner);
+
+                } else if (minVerticalDistance <= verticalDistance
+                           && verticalDistance <= maxVerticalDistance
+                           && Math.abs(horizontalDistance - Math.min(outer.getLength(), inner.getLength())) < 0.1) {
+
+                    boolean characterOverlap = false;
+                    int overlappingCount = 0;
+                    for (Character outerCharacter : outer.getCharacters()) {
+                        for (Character innerCharacter : inner.getCharacters()) {
+                            double characterOverlapDistance = outerCharacter.overlappingDistance(innerCharacter);
+                            if (characterOverlapDistance > 2) {
+                                characterOverlap = true;
+                            }
+                            if (characterOverlapDistance > 0) {
+                                overlappingCount++;
+                            }
+                        }
+                    }
+                    if (!characterOverlap && overlappingCount <= 2) {
+                        unionFind.union(outer, inner);
+                    }
+                }
+
+            });
+        });
+
+        List<Line> outputZone = new ArrayList<>();
+        for (Set<Line> group : unionFind.getGroups()) {
+            List<Character> characters = new ArrayList<>();
+            for (Line line : group) {
+                characters.addAll(line.getCharacters());
+            }
+            characters.sort(Comparator.comparingDouble(Character::getX));
+
+            outputZone.add(new Line(characters, characterSpacing));
+        }
+
+        return new Zone(outputZone.stream()
+                                .sorted(Comparator.comparing(Line::getY0))
+                                .collect(Collectors.toList()));
     }
 
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/docstrum/utils/DoubleUtils.java

@@ -7,12 +7,9 @@ public class DoubleUtils {
         if (Double.isNaN(d1) || Double.isNaN(d2)) {
             return Double.compare(d1, d2);
         }
-
-        if (Math.abs(d1 - d2) < precision) {
-            return 0;
-        }
-
-        return Double.compare(d1, d2);
+        long i1 = Math.round(d1 / (precision == 0 ? 1 : precision));
+        long i2 = Math.round(d2 / (precision == 0 ? 1 : precision));
+        return Long.compare(i1, i2);
     }
 
 }

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

@@ -1,93 +1,73 @@
-package com.knecon.fforesight.service.layoutparser.processor.services.mapper;
+package com.knecon.fforesight.service.layoutparser.processor.markdown;
 
 import java.util.ArrayList;
 import java.util.HashSet;
 import java.util.LinkedList;
 import java.util.List;
 import java.util.Map;
+import java.util.Objects;
 import java.util.Set;
 import java.util.TreeMap;
+import java.util.stream.Stream;
 
-import org.commonmark.Extension;
 import org.commonmark.ext.gfm.tables.TableBlock;
 import org.commonmark.ext.gfm.tables.TableBody;
 import org.commonmark.ext.gfm.tables.TableCell;
 import org.commonmark.ext.gfm.tables.TableHead;
 import org.commonmark.ext.gfm.tables.TableRow;
-import org.commonmark.ext.gfm.tables.TablesExtension;
-import org.commonmark.node.Block;
 import org.commonmark.node.Document;
 import org.commonmark.node.Emphasis;
 import org.commonmark.node.HardLineBreak;
 import org.commonmark.node.Heading;
 import org.commonmark.node.Node;
-import org.commonmark.node.Paragraph;
+import org.commonmark.node.SoftLineBreak;
 import org.commonmark.node.StrongEmphasis;
 import org.commonmark.node.Text;
-import org.commonmark.renderer.markdown.MarkdownRenderer;
 
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.AbstractNodeVisitor;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.TextRange;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Headline;
+import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Paragraph;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.SemanticNode;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Table;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.TextBlock;
-import com.knecon.fforesight.service.layoutparser.processor.model.text.FontStyle;
 
-public class MarkdownMapper extends AbstractNodeVisitor {
+import lombok.experimental.UtilityClass;
 
-    Document markdownDocument = new Document();
+@UtilityClass
+public class DocumentDataParser {
 
+    public Document parse(Stream<SemanticNode> semanticNodes) {
 
-    public String toMarkdownContent(com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Document document) {
-
-        visit(document);
-
-        return buildRenderer().render(this.markdownDocument);
+        Document document = new Document();
+        semanticNodes.map(DocumentDataParser::parseNode)
+                .filter(Objects::nonNull)
+                .forEach(document::appendChild);
+        return document;
     }
 
 
-    @Override
-    public void visit(Headline headline) {
+    private Node parseNode(SemanticNode semanticNode) {
 
-        markdownDocument.appendChild(parseHeadline(headline));
+        return switch (semanticNode.getType()) {
+            case HEADLINE -> parseHeadline((Headline) semanticNode);
+            case PARAGRAPH -> parseParagraph((Paragraph) semanticNode);
+            case TABLE -> parseTable((Table) semanticNode);
+            default -> null;
+        };
     }
 
 
-    @Override
-    public void visit(com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Paragraph paragraph) {
-
-        markdownDocument.appendChild(parseParagraph(paragraph));
-    }
-
-
-    @Override
-    public void visit(Table table) {
-
-        markdownDocument.appendChild(parseTable(table));
-    }
-
-
-    private static MarkdownRenderer buildRenderer() {
-
-        List<Extension> extensions = List.of(TablesExtension.create());
-        return MarkdownRenderer.builder().extensions(extensions).build();
-    }
-
-
-    private Block parseTable(Table table) {
-
-//        if (table.getNumberOfRows() == 1 && table.getNumberOfCols() == 1) {
-//            org.commonmark.node.Paragraph markdownParagraph = new org.commonmark.node.Paragraph();
-//            parseTextBlock(table.getTextBlock(), true).forEach(markdownParagraph::appendChild);
-//            return markdownParagraph;
-//        }
+    private TableBlock parseTable(Table table) {
 
         TableBlock tableNode = new TableBlock();
         TableHead head = new TableHead();
         TableRow tableRow = createTableRow(table, 0);
         head.appendChild(tableRow);
         int row = 1;
+        for (; row < table.getNumberOfRows() && table.streamRow(row)
+                .allMatch(com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.TableCell::isHeader); row++) {
+            head.appendChild(createTableRow(table, row));
+        }
         tableNode.appendChild(head);
         TableBody tableBody = new TableBody();
         for (; row < table.getNumberOfRows(); row++) {
@@ -102,7 +82,7 @@ public class MarkdownMapper extends AbstractNodeVisitor {
 
         TableRow tableRow = new TableRow();
         table.streamRow(row)
-                .map(this::createTableCell)
+                .map(DocumentDataParser::createTableCell)
                 .forEach(tableRow::appendChild);
         return tableRow;
     }
@@ -111,22 +91,16 @@ public class MarkdownMapper extends AbstractNodeVisitor {
     private Node createTableCell(com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.TableCell tc) {
 
         var cell = new TableCell();
-        List<SemanticNode> childNodes = tc.streamChildren()
-                .toList();
-        if (childNodes.isEmpty()) {
-            parseTextBlock(tc.getTextBlock(), false).forEach(cell::appendChild);
-        } else {
-            childNodes.forEach(semanticNode -> parseTextBlock(semanticNode.getTextBlock(), false).forEach(cell::appendChild));
-        }
+        parseTextBlock(tc.getTextBlock()).forEach(cell::appendChild);
         return cell;
     }
 
 
-    private Paragraph parseParagraph(com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Paragraph paragraph) {
+    private org.commonmark.node.Paragraph parseParagraph(Paragraph paragraph) {
 
-        org.commonmark.node.Paragraph markdownParagraph = new org.commonmark.node.Paragraph();
-        parseTextBlock(paragraph.getTextBlock(), true).forEach(markdownParagraph::appendChild);
-        return markdownParagraph;
+        org.commonmark.node.Paragraph heading = new org.commonmark.node.Paragraph();
+        parseTextBlockWithLineBreaks(paragraph.getTextBlock()).forEach(heading::appendChild);
+        return heading;
     }
 
 
@@ -134,90 +108,86 @@ public class MarkdownMapper extends AbstractNodeVisitor {
 
         Heading heading = new Heading();
         heading.setLevel(headline.getTreeId().size());
-        heading.appendChild(parseTextBlockAsText(headline.getTextBlock()));
+        parseTextBlockWithLineBreaks(headline.getTextBlock()).forEach(heading::appendChild);
         return heading;
 
     }
 
-
-    private Text parseTextBlockAsText(TextBlock textBlock) {
-
-        return new Text(textBlock.getSearchText());
-    }
-
-
-    private List<Node> parseTextBlock(TextBlock textBlock, boolean includeLineBreaks) {
+    private List<Node> parseTextBlockWithLineBreaks(TextBlock textBlock) {
 
         LinkedList<Node> result = new LinkedList<>();
         List<TextRangeWithTextType> textRanges = mergeTextStyles(textBlock);
-
-        String fullText = getText(textBlock, textBlock.getTextRange(), includeLineBreaks);
-        List<Integer> lineTextSizes = getLineTextSizes(fullText);
-        int idx = 0;
-        int charCount = 0;
         for (TextRangeWithTextType textRange : textRanges) {
-            String text = getText(textBlock, textRange.textRange(), includeLineBreaks);
+            if (textBlock.subSequenceWithLineBreaks(textRange.textRange()).equals("\n")) {
+                result.add(new HardLineBreak());
+            }
+            String text = textBlock.subSequenceWithLineBreaks(textRange.textRange());
             String[] lines = text.split("\n");
             for (String line : lines) {
-                charCount += line.length();
+                String cleanedLine = line.trim();
+                if (cleanedLine.isEmpty()) {
+                    result.add(new HardLineBreak());
+                    continue;
+                }
                 switch (textRange.fontStyle()) {
-                    case REGULAR -> result.add(new Text(line));
+                    case REGULAR -> result.add(new Text(cleanedLine));
                     case BOLD -> {
                         StrongEmphasis boldBlock = new StrongEmphasis();
-                        boldBlock.appendChild(new Text(line));
+                        boldBlock.appendChild(new Text(cleanedLine));
                         result.add(boldBlock);
                     }
                     case ITALIC -> {
-                        Emphasis italicBlock = new Emphasis();
-                        italicBlock.appendChild(new Text(line));
+                        Emphasis italicBlock = new Emphasis("_");
+                        italicBlock.appendChild(new Text(cleanedLine));
                         result.add(italicBlock);
                     }
                     case BOLD_ITALIC -> {
-                        Emphasis italicBlock = new Emphasis();
+                        Emphasis italicBlock = new Emphasis("_");
 
                         StrongEmphasis boldBlock = new StrongEmphasis();
-                        boldBlock.appendChild(new Text(line));
+                        boldBlock.appendChild(new Text(cleanedLine));
 
                         italicBlock.appendChild(boldBlock);
                         result.add(italicBlock);
                     }
                 }
-                if (includeLineBreaks && lineTextSizes.get(idx).equals(charCount)) {
-                    result.add(new HardLineBreak());
-                    idx++;
-                }
-
+                result.add(new HardLineBreak());
             }
         }
-        if (!result.isEmpty() && result.getLast() instanceof HardLineBreak) {
-            result.removeLast();
-        }
+        result.removeLast();
         return result;
     }
 
 
-    private static List<Integer> getLineTextSizes(String input) {
+    private List<Node> parseTextBlock(TextBlock textBlock) {
 
-        String[] parts = input.split("\n");
-        List<Integer> textSizes = new ArrayList<>();
+        List<Node> result = new ArrayList<>();
+        List<TextRangeWithTextType> textRanges = mergeTextStyles(textBlock);
+        for (TextRangeWithTextType textRange : textRanges) {
+            switch (textRange.fontStyle()) {
+                case REGULAR -> result.add(new Text(textBlock.subSequence(textRange.textRange()).toString()));
+                case BOLD -> {
+                    StrongEmphasis boldBlock = new StrongEmphasis();
+                    boldBlock.appendChild(new Text(textBlock.subSequence(textRange.textRange()).toString()));
+                    result.add(boldBlock);
+                }
+                case ITALIC -> {
+                    Emphasis italicBlock = new Emphasis("_");
+                    italicBlock.appendChild(new Text(textBlock.subSequence(textRange.textRange()).toString()));
+                    result.add(italicBlock);
+                }
+                case BOLD_ITALIC -> {
+                    Emphasis italicBlock = new Emphasis("_");
 
-        int size = 0;
-        for (int i = 0; i < parts.length; i++) {
-            size += parts[i].length();
-            textSizes.add(size);
+                    StrongEmphasis boldBlock = new StrongEmphasis();
+                    boldBlock.appendChild(new Text(textBlock.subSequence(textRange.textRange()).toString()));
+
+                    italicBlock.appendChild(boldBlock);
+                    result.add(italicBlock);
+                }
+            }
         }
-
-        if (textSizes.isEmpty()) {
-            textSizes.add(0);
-        }
-
-        return textSizes;
-    }
-
-
-    private static String getText(TextBlock textBlock, TextRange textRange, boolean includeLineBreaks) {
-
-        return includeLineBreaks ? textBlock.subSequenceWithLineBreaks(textRange) : textBlock.subSequence(textRange).toString();
+        return result;
     }
 
 
@@ -280,7 +250,9 @@ public class MarkdownMapper extends AbstractNodeVisitor {
             result.add(new TextRangeWithTextType(new TextRange(start, textBlock.getTextRange().end()), style));
         }
 
-        return result;
+        return result.stream()
+                .filter(t -> t.textRange.length() > 1)
+                .toList();
     }
 
 
@@ -298,6 +270,12 @@ public class MarkdownMapper extends AbstractNodeVisitor {
     }
 
 
+    enum FontStyle {
+        REGULAR,
+        BOLD,
+        ITALIC,
+        BOLD_ITALIC;
+    }
 
     record FontStyleChange(boolean enter, FontStyle style) {
 
@@ -323,4 +301,5 @@ public class MarkdownMapper extends AbstractNodeVisitor {
     record TextRangeWithTextType(TextRange textRange, FontStyle fontStyle) {
 
     }
+
 }

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

@@ -0,0 +1,13 @@
+package com.knecon.fforesight.service.layoutparser.processor.markdown;
+
+import org.commonmark.node.IndentedCodeBlock;
+import org.commonmark.node.Paragraph;
+
+import com.knecon.fforesight.service.layoutparser.processor.utils.TokenCounter;
+
+public class ExtraTokens {
+
+    public static int INDENTED_CODE_BLOCK = 10;
+    public static int PARAGRAPH = 10;
+
+}

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

@@ -0,0 +1,527 @@
+package com.knecon.fforesight.service.layoutparser.processor.markdown;
+
+import static com.knecon.fforesight.service.layoutparser.processor.MarkdownParsingPipeline.buildRenderer;
+import static com.knecon.fforesight.service.layoutparser.processor.utils.TokenCounter.countTokens;
+
+import java.text.BreakIterator;
+import java.util.Collections;
+import java.util.Deque;
+import java.util.LinkedList;
+import java.util.List;
+import java.util.Locale;
+import java.util.stream.Collectors;
+
+import org.commonmark.ext.gfm.tables.TableBlock;
+import org.commonmark.ext.gfm.tables.TableBody;
+import org.commonmark.node.AbstractVisitor;
+import org.commonmark.node.BlockQuote;
+import org.commonmark.node.BulletList;
+import org.commonmark.node.Code;
+import org.commonmark.node.CustomBlock;
+import org.commonmark.node.CustomNode;
+import org.commonmark.node.Document;
+import org.commonmark.node.Emphasis;
+import org.commonmark.node.FencedCodeBlock;
+import org.commonmark.node.HardLineBreak;
+import org.commonmark.node.Heading;
+import org.commonmark.node.HtmlBlock;
+import org.commonmark.node.HtmlInline;
+import org.commonmark.node.Image;
+import org.commonmark.node.IndentedCodeBlock;
+import org.commonmark.node.Link;
+import org.commonmark.node.LinkReferenceDefinition;
+import org.commonmark.node.ListBlock;
+import org.commonmark.node.ListItem;
+import org.commonmark.node.Node;
+import org.commonmark.node.OrderedList;
+import org.commonmark.node.Paragraph;
+import org.commonmark.node.SoftLineBreak;
+import org.commonmark.node.StrongEmphasis;
+import org.commonmark.node.Text;
+import org.commonmark.node.ThematicBreak;
+import org.commonmark.renderer.Renderer;
+
+import com.knecon.fforesight.service.layoutparser.processor.utils.TokenCounter;
+
+import lombok.AccessLevel;
+import lombok.experimental.FieldDefaults;
+import lombok.extern.slf4j.Slf4j;
+
+@Slf4j
+@FieldDefaults(level = AccessLevel.PRIVATE)
+public class MarkdownChunker extends AbstractVisitor {
+
+    NodeCopier nodeCopier = NodeCopier.INSTANCE;
+
+    final int tokenLimit;
+    List<Document> allChunks;
+
+    Deque<Heading> currentHeadings;
+    Document chunk;
+    boolean validChunk;
+
+
+    public MarkdownChunker(int tokenLimit) {
+
+        this.tokenLimit = tokenLimit;
+        allChunks = new LinkedList<>();
+        currentHeadings = new LinkedList<>();
+        startNewChunk();
+    }
+
+
+    public List<Document> getResult() {
+
+        for (Document chunk : allChunks) {
+            if (countTokens(chunk) > tokenLimit) {
+                throwUnsplittableNodeError(chunk);
+            }
+        }
+        return allChunks;
+    }
+
+
+    @Override
+    public void visit(Heading heading) {
+
+        if (heading.getLevel() > 4) {
+            addToChunk(heading);
+        }
+        if (currentHeadings.isEmpty() || currentHeadings.peek().getLevel() < heading.getLevel()) {
+            currentHeadings.push(heading);
+        } else {
+            while (!currentHeadings.isEmpty() && currentHeadings.peek().getLevel() >= heading.getLevel()) {
+                currentHeadings.pop();
+            }
+            currentHeadings.push(heading);
+        }
+
+        startNewChunk();
+    }
+
+
+    private void startNewChunk() {
+
+        if (!validChunk && !allChunks.isEmpty()) {
+            allChunks.remove(allChunks.size() - 1);
+        }
+        validChunk = false;
+        chunk = buildNewChunk();
+        allChunks.add(chunk);
+    }
+
+
+    private Document buildNewChunk() {
+
+        Document document = new Document();
+        List<Node> headingCopies = currentHeadings.stream()
+                .map(nodeCopier::copyNodeWithChildren)
+                .collect(Collectors.toList());
+        Collections.reverse(headingCopies);
+        headingCopies.forEach(document::appendChild);
+        return document;
+    }
+
+
+    public int currentTokenCount() {
+
+        return TokenCounter.countTokens(chunk);
+    }
+
+
+    public boolean fitsTokenLimit(Node node) {
+
+        Document document = buildNewChunk();
+        document.appendChild(nodeCopier.copyNodeWithChildren(node));
+        return TokenCounter.countTokens(document) <= tokenLimit;
+    }
+
+
+    private void addToChunk(Node node) {
+
+        chunk.appendChild(node);
+
+        if (currentTokenCount() <= tokenLimit) {
+            return;
+        }
+
+        node.unlink();
+        startNewChunk();
+        chunk.appendChild(node);
+
+        if (currentTokenCount() > tokenLimit) { // node is too large and won't fit in tokenLimit, split is necessary
+            node.unlink();
+            startNewChunk();
+            splitNodeAndAddToChunk(node);
+            return;
+        }
+
+        validChunk = true;
+
+    }
+
+
+    private void splitNodeAndAddToChunk(Node node) {
+
+        if (node instanceof TableBlock tableBlock) {
+            splitTable(tableBlock);
+            return;
+        } else if (node instanceof BulletList bulletList) {
+            splitList(bulletList);
+            return;
+        } else if (node instanceof OrderedList orderedList) {
+            splitList(orderedList);
+            return;
+        } else if (node instanceof Paragraph paragraph) {
+            splitParagraph(paragraph);
+            return;
+        } else if (node instanceof IndentedCodeBlock indentedCodeBlock) {
+            splitCodeBlock(indentedCodeBlock);
+            return;
+        }
+        throwUnsplittableNodeError(node);
+    }
+
+
+    private void splitCodeBlock(IndentedCodeBlock indentedCodeBlock) {
+
+        List<IndentedCodeBlock> splitBlocks = new LinkedList<>();
+        StringBuilder sb = new StringBuilder();
+        BreakIterator lineIterator = BreakIterator.getLineInstance(Locale.ENGLISH);
+        lineIterator.setText(indentedCodeBlock.getLiteral());
+        int start = lineIterator.first();
+        for (int end = lineIterator.next(); end != BreakIterator.DONE; start = end, end = lineIterator.next()) {
+            String sentence = indentedCodeBlock.getLiteral().substring(start, end);
+            if (!fitsTokenLimit(buildIndentedCodeBlock(sb.toString()))) {
+                sb.replace(sb.length() - sentence.length(), sb.length(), "");
+                IndentedCodeBlock block = buildIndentedCodeBlock(sb.toString());
+                splitBlocks.add(block);
+                sb = new StringBuilder();
+            }
+            sb.append(sentence);
+        }
+        if (!sb.isEmpty()) {
+            if (fitsTokenLimit(buildIndentedCodeBlock(sb.toString()))) {
+                splitBlocks.add(buildIndentedCodeBlock(sb.toString()));
+            } else {
+                int mid = sb.length() / 2;
+                splitBlocks.add(buildIndentedCodeBlock(sb.substring(0, mid)));
+                splitBlocks.add(buildIndentedCodeBlock(sb.substring(mid, sb.length())));
+            }
+        }
+
+        splitBlocks.forEach(this::addToChunk);
+    }
+
+
+    private static IndentedCodeBlock buildIndentedCodeBlock(String string) {
+
+        IndentedCodeBlock block = new IndentedCodeBlock();
+        block.setLiteral(string);
+        return block;
+    }
+
+
+    private void splitParagraph(Paragraph paragraph) {
+
+        if (fitsTokenLimit(paragraph)) {
+            addToChunk(paragraph);
+            return;
+        }
+
+        List<Node> children = collectAllChildNodes(paragraph);
+
+        if (children.size() == 1) {
+            if (children.get(0) instanceof Text text) {
+                List<Text> splitTexts = splitText(text);
+                for (Text splitText : splitTexts) {
+                    Paragraph paragraph1 = new Paragraph();
+                    paragraph1.appendChild(splitText);
+                    addToChunk(paragraph1);
+                }
+                return;
+            }
+            throwUnsplittableNodeError(children.get(0));
+        }
+
+        Paragraph paragraph1 = new Paragraph();
+        Paragraph paragraph2 = new Paragraph();
+
+        int mid = children.size() / 2;
+        children.subList(0, mid)
+                .forEach(paragraph1::appendChild);
+        children.subList(mid, children.size())
+                .forEach(paragraph2::appendChild);
+
+        splitParagraph(paragraph1);
+        splitParagraph(paragraph2);
+    }
+
+
+    private void throwUnsplittableNodeError(Node node) {
+
+        Renderer renderer = buildRenderer();
+        String renderedNode = renderer.render(node);
+        log.error(renderedNode);
+        throw new IllegalArgumentException(String.format("Node %s exceeds token limit (%d/%d) and can't be split!", node, countTokens(renderedNode), tokenLimit));
+    }
+
+
+    private static List<Node> collectAllChildNodes(Node parent) {
+
+        List<Node> children = new LinkedList<>();
+        Node next;
+        for (Node child = parent.getFirstChild(); child != null; child = next) {
+            next = child.getNext();
+            children.add(child);
+        }
+        return children;
+    }
+
+
+    private List<Text> splitText(Text text) {
+
+        List<Text> splitTexts = new LinkedList<>();
+        StringBuilder sb = new StringBuilder();
+        BreakIterator sentenceIterator = BreakIterator.getSentenceInstance(Locale.ENGLISH);
+        sentenceIterator.setText(text.getLiteral());
+        int start = sentenceIterator.first();
+        for (int end = sentenceIterator.next(); end != BreakIterator.DONE; start = end, end = sentenceIterator.next()) {
+            String sentence = text.getLiteral().substring(start, end);
+            if (!fitsTokenLimit(buildParagraphWithText(sb))) {
+                sb.replace(sb.length() - sentence.length(), sb.length(), "");
+                splitTexts.add(new Text(sb.toString()));
+                sb = new StringBuilder();
+            }
+            sb.append(sentence);
+        }
+        if (!sb.isEmpty()) {
+            if (fitsTokenLimit(buildParagraphWithText(sb))) {
+                splitTexts.add(new Text(sb.toString()));
+            } else {
+                int mid = sb.length() / 2;
+                splitTexts.add(new Text(sb.substring(0, mid)));
+                splitTexts.add(new Text(sb.substring(mid, sb.length())));
+            }
+        }
+        return splitTexts;
+    }
+
+
+    private static Paragraph buildParagraphWithText(StringBuilder sb) {
+
+        Paragraph paragraph = new Paragraph();
+        paragraph.appendChild(new Text(sb.toString()));
+        return paragraph;
+    }
+
+
+    private void splitList(BulletList bulletList) {
+
+        if (fitsTokenLimit(bulletList)) {
+            addToChunk(bulletList);
+            return;
+        }
+
+        BulletList list1 = new BulletList();
+        BulletList list2 = new BulletList();
+
+        splitLists(bulletList, list1, list2);
+    }
+
+
+    private void splitList(OrderedList orderedList) {
+
+        if (fitsTokenLimit(orderedList)) {
+            addToChunk(orderedList);
+            return;
+        }
+
+        OrderedList list1 = new OrderedList();
+        OrderedList list2 = new OrderedList();
+
+        splitLists(orderedList, list1, list2);
+    }
+
+
+    private void splitLists(ListBlock originList, ListBlock list1, ListBlock list2) {
+
+        List<Node> listItems = collectAllChildNodes(originList);
+
+        if (listItems.size() == 1) {
+            collectAllChildNodes(listItems.get(0)).forEach(this::addToChunk);
+        }
+
+        int mid = listItems.size() / 2;
+        listItems.subList(0, mid)
+                .forEach(list1::appendChild);
+        listItems.subList(mid, listItems.size())
+                .forEach(list2::appendChild);
+
+        splitNodeAndAddToChunk(list1);
+        splitNodeAndAddToChunk(list2);
+    }
+
+
+    private void splitTable(TableBlock tableBlock) {
+
+        if (fitsTokenLimit(tableBlock)) {
+            addToChunk(tableBlock);
+            return;
+        }
+
+        TableBlock tableBlock1 = new TableBlock();
+        TableBlock tableBlock2 = new TableBlock();
+
+        tableBlock1.appendChild(nodeCopier.copy(tableBlock.getFirstChild()));
+        tableBlock2.appendChild(nodeCopier.copy(tableBlock.getFirstChild()));
+
+        TableBody tableBody1 = new TableBody();
+        TableBody tableBody2 = new TableBody();
+
+        List<Node> tableRows = collectAllChildNodes(tableBlock.getLastChild());
+
+        if (tableRows.isEmpty()) {
+            throw new IllegalArgumentException("The table headers already exceeds the token limit");
+        }
+        if (tableRows.size() == 1) {
+            throw new IllegalArgumentException("A single table row already exceeds the token limit");
+        }
+
+        int mid = tableRows.size() / 2;
+        tableRows.subList(0, mid)
+                .forEach(tableBody1::appendChild);
+        tableRows.subList(mid, tableRows.size())
+                .forEach(tableBody2::appendChild);
+
+        splitTable(tableBlock1);
+        splitTable(tableBlock2);
+    }
+
+
+    public void visit(BlockQuote blockQuote) {
+
+        this.addToChunk(blockQuote);
+    }
+
+
+    public void visit(BulletList bulletList) {
+
+        this.addToChunk(bulletList);
+    }
+
+
+    public void visit(Code code) {
+
+        this.addToChunk(code);
+    }
+
+
+    public void visit(Emphasis emphasis) {
+
+        this.addToChunk(emphasis);
+    }
+
+
+    public void visit(FencedCodeBlock fencedCodeBlock) {
+
+        this.addToChunk(fencedCodeBlock);
+    }
+
+
+    public void visit(HardLineBreak hardLineBreak) {
+
+        this.addToChunk(hardLineBreak);
+    }
+
+
+    public void visit(ThematicBreak thematicBreak) {
+
+        this.addToChunk(thematicBreak);
+    }
+
+
+    public void visit(HtmlInline htmlInline) {
+
+        this.addToChunk(htmlInline);
+    }
+
+
+    public void visit(HtmlBlock htmlBlock) {
+
+        this.addToChunk(htmlBlock);
+    }
+
+
+    public void visit(Image image) {
+
+        this.addToChunk(image);
+    }
+
+
+    public void visit(IndentedCodeBlock indentedCodeBlock) {
+
+        this.addToChunk(indentedCodeBlock);
+    }
+
+
+    public void visit(Link link) {
+
+        this.addToChunk(link);
+    }
+
+
+    public void visit(ListItem listItem) {
+
+        this.addToChunk(listItem);
+    }
+
+
+    public void visit(OrderedList orderedList) {
+
+        this.addToChunk(orderedList);
+    }
+
+
+    public void visit(Paragraph paragraph) {
+
+        this.addToChunk(paragraph);
+    }
+
+
+    public void visit(SoftLineBreak softLineBreak) {
+
+        this.addToChunk(softLineBreak);
+    }
+
+
+    public void visit(StrongEmphasis strongEmphasis) {
+
+        this.addToChunk(strongEmphasis);
+    }
+
+
+    public void visit(Text text) {
+
+        this.addToChunk(text);
+    }
+
+
+    public void visit(LinkReferenceDefinition linkReferenceDefinition) {
+
+        this.addToChunk(linkReferenceDefinition);
+    }
+
+
+    public void visit(CustomBlock customBlock) {
+
+        this.addToChunk(customBlock);
+    }
+
+
+    public void visit(CustomNode customNode) {
+
+        this.addToChunk(customNode);
+    }
+
+}

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

@@ -0,0 +1,171 @@
+package com.knecon.fforesight.service.layoutparser.processor.markdown;
+
+import org.commonmark.ext.gfm.tables.TableBlock;
+import org.commonmark.ext.gfm.tables.TableBody;
+import org.commonmark.ext.gfm.tables.TableCell;
+import org.commonmark.ext.gfm.tables.TableHead;
+import org.commonmark.ext.gfm.tables.TableRow;
+import org.commonmark.node.BlockQuote;
+import org.commonmark.node.BulletList;
+import org.commonmark.node.Code;
+import org.commonmark.node.Document;
+import org.commonmark.node.Emphasis;
+import org.commonmark.node.FencedCodeBlock;
+import org.commonmark.node.HardLineBreak;
+import org.commonmark.node.Heading;
+import org.commonmark.node.HtmlBlock;
+import org.commonmark.node.HtmlInline;
+import org.commonmark.node.Image;
+import org.commonmark.node.IndentedCodeBlock;
+import org.commonmark.node.Link;
+import org.commonmark.node.LinkReferenceDefinition;
+import org.commonmark.node.ListItem;
+import org.commonmark.node.Node;
+import org.commonmark.node.OrderedList;
+import org.commonmark.node.Paragraph;
+import org.commonmark.node.SoftLineBreak;
+import org.commonmark.node.StrongEmphasis;
+import org.commonmark.node.Text;
+import org.commonmark.node.ThematicBreak;
+import org.mapstruct.Mapper;
+import org.mapstruct.Mapping;
+import org.mapstruct.factory.Mappers;
+
+@Mapper
+public interface NodeCopier {
+
+    NodeCopier INSTANCE = Mappers.getMapper(NodeCopier.class);
+
+
+    default Node copyNodeWithChildren(Node node) {
+
+        Node copy = copy(node);
+
+        Node next;
+        for (Node child = node.getFirstChild(); child != null; child = next) {
+            next = child.getNext();
+            copy.appendChild(copyNodeWithChildren(child));
+        }
+        return copy;
+    }
+
+
+    default Node copy(Node node) {
+
+        return switch (node.getClass().getSimpleName()) {
+            case "BlockQuote" -> copy((BlockQuote) node);
+            case "BulletList" -> copy((BulletList) node);
+            case "Code" -> copy((Code) node);
+            case "Document" -> copy((Document) node);
+            case "Emphasis" -> copy((Emphasis) node);
+            case "FencedCodeBlock" -> copy((FencedCodeBlock) node);
+            case "HardLineBreak" -> copy((HardLineBreak) node);
+            case "Heading" -> copy((Heading) node);
+            case "HtmlBlock" -> copy((HtmlBlock) node);
+            case "HtmlInline" -> copy((HtmlInline) node);
+            case "Image" -> copy((Image) node);
+            case "IndentedCodeBlock" -> copy((IndentedCodeBlock) node);
+            case "Link" -> copy((Link) node);
+            case "LinkReferenceDefinition" -> copy((LinkReferenceDefinition) node);
+            case "ListItem" -> copy((ListItem) node);
+            case "OrderedList" -> copy((OrderedList) node);
+            case "Paragraph" -> copy((Paragraph) node);
+            case "SoftLineBreak" -> copy((SoftLineBreak) node);
+            case "StrongEmphasis" -> copy((StrongEmphasis) node);
+            case "Text" -> copy((Text) node);
+            case "ThematicBreak" -> copy((ThematicBreak) node);
+            case "TableBlock" -> copy((TableBlock) node);
+            case "TableBody" -> copy((TableBody) node);
+            case "TableCell" -> copy((TableCell) node);
+            case "TableHead" -> copy((TableHead) node);
+            case "TableRow" -> copy((TableRow) node);
+            default -> throw new IllegalArgumentException("No copy method found for class: " + node.getClass().getName());
+        };
+    }
+
+
+    BlockQuote copy(BlockQuote blockQuote);
+
+
+    @Mapping(target = "bulletMarker", ignore = true)
+    BulletList copy(BulletList bulletList);
+
+
+    Code copy(Code code);
+
+
+    Document copy(Document document);
+
+
+    @Mapping(target = "delimiter", source = "openingDelimiter")
+    Emphasis copy(Emphasis emphasis);
+
+
+    @Mapping(target = "fenceChar", ignore = true)
+    @Mapping(target = "fenceLength", ignore = true)
+    FencedCodeBlock copy(FencedCodeBlock fencedCodeBlock);
+
+
+    HardLineBreak copy(HardLineBreak hardLineBreak);
+
+
+    Heading copy(Heading heading);
+
+
+    HtmlBlock copy(HtmlBlock htmlBlock);
+
+
+    HtmlInline copy(HtmlInline htmlInline);
+
+
+    Image copy(Image image);
+
+
+    IndentedCodeBlock copy(IndentedCodeBlock indentedCodeBlock);
+
+
+    Link copy(Link link);
+
+
+    LinkReferenceDefinition copy(LinkReferenceDefinition linkReferenceDefinition);
+
+
+    ListItem copy(ListItem listItem);
+
+
+    @Mapping(target = "startNumber", ignore = true)
+    @Mapping(target = "delimiter", ignore = true)
+    OrderedList copy(OrderedList orderedList);
+
+
+    Paragraph copy(Paragraph paragraph);
+
+
+    SoftLineBreak copy(SoftLineBreak softLineBreak);
+
+
+    @Mapping(target = "delimiter", source = "openingDelimiter")
+    StrongEmphasis copy(StrongEmphasis strongEmphasis);
+
+
+    Text copy(Text text);
+
+
+    ThematicBreak copy(ThematicBreak thematicBreak);
+
+
+    TableBlock copy(TableBlock tableBlock);
+
+
+    TableBody copy(TableBody tableBody);
+
+
+    TableCell copy(TableCell tableCell);
+
+
+    TableHead copy(TableHead tableHead);
+
+
+    TableRow copy(TableRow tableRow);
+
+}

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

@@ -0,0 +1,65 @@
+package com.knecon.fforesight.service.layoutparser.processor.markdown;
+
+import java.lang.reflect.Field;
+
+import org.commonmark.node.Node;
+
+public class ReflectionNodeCopier {
+
+    NodeCopier mapperNodeCopier;
+
+
+
+    public static Node copyNode(Node node) {
+
+        Node copy = deepCopy(node);
+        copyChildren(node, copy);
+        return copy;
+    }
+
+
+    private static void copyChildren(Node nodeToCopy, Node copy) {
+
+        Node next;
+        for (Node node = nodeToCopy.getFirstChild(); node != null; node = next) {
+            next = node.getNext();
+            copy.appendChild(copyNode(node));
+        }
+    }
+
+
+    private static <T> T deepCopy(T object) {
+
+        try {
+            Class<?> clazz = object.getClass();
+            T copy = (T) clazz.getDeclaredConstructor().newInstance();
+
+            for (Field field : clazz.getDeclaredFields()) {
+                field.setAccessible(true);
+                Object value = field.get(object);
+                if (isPrimitiveOrWrapper(field.getType()) || field.getType().equals(String.class)) {
+                    field.set(copy, value);
+                }
+            }
+            return copy;
+        } catch (Exception e) {
+            e.printStackTrace();
+            return null;
+        }
+    }
+
+
+    private static boolean isPrimitiveOrWrapper(Class<?> type) {
+
+        return type.isPrimitive()
+               || type == Boolean.class
+               || type == Byte.class
+               || type == Character.class
+               || type == Double.class
+               || type == Float.class
+               || type == Integer.class
+               || type == Long.class
+               || type == Short.class;
+    }
+
+}

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

@@ -4,7 +4,7 @@ import java.util.HashSet;
 import java.util.Set;
 
 import com.fasterxml.jackson.annotation.JsonIgnore;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.LayoutEngineProto.LayoutEngine;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.LayoutEngine;
 import com.knecon.fforesight.service.layoutparser.processor.docstrum.model.BoundingBox;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
 

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

@@ -7,7 +7,7 @@ import com.knecon.fforesight.service.layoutparser.processor.model.outline.Outlin
 import com.knecon.fforesight.service.layoutparser.processor.model.outline.TableOfContents;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.StringFrequencyCounter;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.UnclassifiedText;
-import com.knecon.fforesight.service.layoutparser.processor.visualization.LayoutDebugLayer;
+import com.knecon.fforesight.service.layoutparser.processor.visualization.LayoutparsingVisualizations;
 
 import lombok.Data;
 import lombok.NoArgsConstructor;
@@ -25,7 +25,7 @@ public class ClassificationDocument {
     private FloatFrequencyCounter fontSizeCounter = new FloatFrequencyCounter();
     private StringFrequencyCounter fontCounter = new StringFrequencyCounter();
     private StringFrequencyCounter fontStyleCounter = new StringFrequencyCounter();
-    private LayoutDebugLayer layoutDebugLayer = new LayoutDebugLayer();
+    private LayoutparsingVisualizations visualizations = new LayoutparsingVisualizations();
     private boolean headlines;
 
     private long rulesVersion;

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

@@ -18,7 +18,6 @@ import lombok.RequiredArgsConstructor;
 
 @Data
 @RequiredArgsConstructor
-
 public class ClassificationPage {
 
     @NonNull
@@ -26,7 +25,7 @@ public class ClassificationPage {
 
     private List<OutlineObject> outlineObjects = new ArrayList<>();
 
-    private List<AbstractPageBlock> headlines = new ArrayList<>();
+    private List<AbstractPageBlock> headlines  = new ArrayList<>();
 
     private List<ClassifiedImage> images = new ArrayList<>();
 
@@ -45,7 +44,7 @@ public class ClassificationPage {
     private float pageWidth;
     private float pageHeight;
 
-    private CleanRulings cleanRulings;
+    CleanRulings cleanRulings;
 
     private Map<String, List<Rectangle2D>> markedContentBboxPerType = new HashMap<>();
 

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

@@ -1,24 +1,23 @@
 package com.knecon.fforesight.service.layoutparser.processor.model;
 
+import java.util.ArrayList;
 import java.util.Collections;
 import java.util.HashMap;
 import java.util.List;
 import java.util.Map;
 import java.util.stream.Collectors;
 
+import com.google.common.base.Functions;
+
 import lombok.Getter;
 
 @Getter
 public class FloatFrequencyCounter {
 
-    Map<Double, Integer> countPerValue = new HashMap<>();
-    boolean changed;
-    Double mostPopularCache;
+    Map<Float, Integer> countPerValue = new HashMap<>();
 
 
-    public void add(double value) {
-
-        changed = true;
+    public void add(float value) {
 
         if (!countPerValue.containsKey(value)) {
             countPerValue.put(value, 1);
@@ -28,11 +27,9 @@ public class FloatFrequencyCounter {
     }
 
 
-    public void addAll(Map<Double, Integer> otherCounter) {
+    public void addAll(Map<Float, Integer> otherCounter) {
 
-        changed = true;
-
-        for (Map.Entry<Double, Integer> entry : otherCounter.entrySet()) {
+        for (Map.Entry<Float, Integer> entry : otherCounter.entrySet()) {
             if (countPerValue.containsKey(entry.getKey())) {
                 countPerValue.put(entry.getKey(), countPerValue.get(entry.getKey()) + entry.getValue());
             } else {
@@ -42,36 +39,38 @@ public class FloatFrequencyCounter {
     }
 
 
-    public Double getMostPopular() {
+    public Float getMostPopular() {
 
-        if (changed || mostPopularCache == null) {
-            Map.Entry<Double, Integer> mostPopular = null;
-            for (Map.Entry<Double, Integer> entry : countPerValue.entrySet()) {
-                if (mostPopular == null || entry.getValue() >= mostPopular.getValue()) {
-                    mostPopular = entry;
-                }
+        Map.Entry<Float, Integer> mostPopular = null;
+        for (Map.Entry<Float, Integer> entry : countPerValue.entrySet()) {
+            if (mostPopular == null || entry.getValue() >= mostPopular.getValue()) {
+                mostPopular = entry;
             }
-            mostPopularCache = mostPopular != null ? mostPopular.getKey() : 0;
-            changed = false;
         }
-
-        return mostPopularCache;
+        return mostPopular != null ? mostPopular.getKey() : null;
     }
 
 
-    public List<Double> getValuesInReverseOrder() {
+    public List<Float> getHighterThanMostPopular() {
 
-        return countPerValue.keySet()
-                .stream()
+        Float mostPopular = getMostPopular();
+        List<Float> higher = new ArrayList<>();
+        for (Float value : countPerValue.keySet()) {
+            if (value > mostPopular) {
+                higher.add(value);
+            }
+        }
+
+        return higher.stream()
                 .sorted(Collections.reverseOrder())
                 .collect(Collectors.toList());
     }
 
 
-    public Double getHighest() {
+    public Float getHighest() {
 
-        Double highest = null;
-        for (Double value : countPerValue.keySet()) {
+        Float highest = null;
+        for (Float value : countPerValue.keySet()) {
             if (highest == null || value > highest) {
                 highest = value;
             }
@@ -79,4 +78,16 @@ public class FloatFrequencyCounter {
         return highest;
     }
 
+
+    public double getAverage() {
+
+        double sum = countPerValue.keySet()
+                .stream()
+                .mapToDouble(fontSize -> fontSize * countPerValue.get(fontSize)).sum();
+        double count = countPerValue.values()
+                .stream()
+                .mapToInt(Integer::intValue).sum();
+        return sum / count;
+    }
+
 }

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

@@ -3,7 +3,7 @@ package com.knecon.fforesight.service.layoutparser.processor.model;
 import java.awt.geom.Rectangle2D;
 import java.util.List;
 
-import com.knecon.fforesight.service.layoutparser.processor.model.text.Word;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPositionSequence;
 
 import lombok.AccessLevel;
 import lombok.AllArgsConstructor;
@@ -16,8 +16,8 @@ import lombok.experimental.FieldDefaults;
 public class LineInformation {
 
     List<Rectangle2D> lineBBox;
-    List<List<Word>> sequencesByLines;
+    List<List<TextPositionSequence>> sequencesByLines;
     List<List<Rectangle2D>> bBoxWithGapsByLines;
-    List<List<List<Word>>> sequencesWithGapsByLines;
+    List<List<List<TextPositionSequence>>> sequencesWithGapsByLines;
 
 }

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

@@ -9,13 +9,12 @@ public enum PageBlockType {
     H6,
     HEADER,
     FOOTER,
+    TITLE,
     PARAGRAPH,
     PARAGRAPH_BOLD,
     PARAGRAPH_ITALIC,
     PARAGRAPH_UNKNOWN,
     OTHER,
-    TABLE_OF_CONTENTS_ITEM,
-    LIST_ITEM,
     TABLE;
 
 

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

@@ -4,7 +4,7 @@ import java.awt.geom.Rectangle2D;
 import java.util.List;
 
 import com.knecon.fforesight.service.layoutparser.processor.model.table.Ruling;
-import com.knecon.fforesight.service.layoutparser.processor.model.text.Word;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPositionSequence;
 
 import lombok.AllArgsConstructor;
 import lombok.Builder;
@@ -15,7 +15,7 @@ import lombok.Getter;
 @AllArgsConstructor
 public class PageContents {
 
-    List<Word> sortedWords;
+    List<TextPositionSequence> sortedTextPositionSequences;
     Rectangle2D cropBox;
     Rectangle2D mediaBox;
     List<Ruling> rulings;

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

@@ -3,7 +3,6 @@ package com.knecon.fforesight.service.layoutparser.processor.model;
 import java.util.Collections;
 import java.util.LinkedList;
 import java.util.List;
-import java.util.Locale;
 import java.util.regex.Matcher;
 import java.util.regex.Pattern;
 
@@ -16,13 +15,11 @@ import lombok.experimental.FieldDefaults;
 @FieldDefaults(makeFinal = true, level = AccessLevel.PRIVATE)
 public class SectionIdentifier {
 
-    public static Pattern numericalIdentifierPattern = Pattern.compile("^[\\s]?(\\d{1,2})(?:[\\s.,;](\\d{1,2}))?(?:[\\s.,;](\\d{1,2}))?(?:[\\s.,;](\\d{1,2}))?");
-    public static Pattern alphanumericIdentifierPattern = Pattern.compile("^[\\s]?[A-Za-z][\\s.,;]?(\\d{1,2})[\\s.,;]?(\\d{1,2})?[\\s.,;]?(\\d{1,2})?[\\s.,;]?(\\d{1,2})?[\\s.,;]?");
+    static Pattern numericalIdentifierPattern = Pattern.compile("^[\\s]?(\\d+)[\\s.,;]?(\\d+)?[\\s.,;]?(\\d+)?[\\s.,;]?(\\d+)?");
 
     public enum Format {
         EMPTY,
         NUMERICAL,
-        ALPHANUMERIC,
         DOCUMENT
     }
 
@@ -44,10 +41,6 @@ public class SectionIdentifier {
         if (numericalIdentifierMatcher.find()) {
             return buildNumericalSectionIdentifier(headline, numericalIdentifierMatcher);
         }
-        Matcher alphanumericIdentifierMatcher = alphanumericIdentifierPattern.matcher(headline);
-        if (alphanumericIdentifierMatcher.find()) {
-            return buildAlphanumericSectionIdentifier(headline, alphanumericIdentifierMatcher);
-        }
         // more formats here
         return SectionIdentifier.empty();
     }
@@ -82,36 +75,7 @@ public class SectionIdentifier {
             }
             identifiers.add(Integer.parseInt(numericalIdentifier.trim()));
         }
-        return new SectionIdentifier(Format.NUMERICAL,
-                                     identifierString,
-                                     identifiers.stream()
-                                             .toList(),
-                                     false);
-    }
-
-
-    private static SectionIdentifier buildAlphanumericSectionIdentifier(String headline, Matcher alphanumericIdentifierMatcher) {
-
-        String identifierString = headline.substring(alphanumericIdentifierMatcher.start(), alphanumericIdentifierMatcher.end());
-
-        String alphanumericIdentifier = alphanumericIdentifierMatcher.group(0).substring(0, 1).toUpperCase(Locale.ENGLISH);
-        int mappedCharacterValue = alphanumericIdentifier.charAt(0) - 'A' + 1;
-        List<Integer> identifiers = new LinkedList<>();
-        identifiers.add(mappedCharacterValue);
-
-        for (int i = 1; i <= 3; i++) {
-            String numericalIdentifier = alphanumericIdentifierMatcher.group(i);
-            if (numericalIdentifier == null || numericalIdentifier.equals("0") || numericalIdentifier.isEmpty() || numericalIdentifier.isBlank()) {
-                break;
-            }
-            identifiers.add(Integer.parseInt(numericalIdentifier.trim()));
-        }
-
-        return new SectionIdentifier(Format.ALPHANUMERIC,
-                                     identifierString,
-                                     identifiers.stream()
-                                             .toList(),
-                                     false);
+        return new SectionIdentifier(Format.NUMERICAL, identifierString, identifiers.stream().toList(), false);
     }
 
 
@@ -159,22 +123,4 @@ public class SectionIdentifier {
         return identifierString;
     }
 
-
-    public boolean isEmpty() {
-
-        return this.format.equals(Format.EMPTY);
-    }
-
-
-    public int level() {
-
-        return identifiers.size();
-    }
-
-
-    protected List<Integer> getIdentifiers() {
-
-        return identifiers;
-    }
-
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/AbstractNodeVisitor.java

@@ -1,94 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.processor.model.graph;
-
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Document;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Footer;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Header;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Headline;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Image;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Paragraph;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Section;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.SemanticNode;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.SuperSection;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Table;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.TableCell;
-
-public abstract class AbstractNodeVisitor implements NodeVisitor {
-
-    @Override
-    public void visit(Document document) {
-
-        visitChildren(document);
-    }
-
-
-    @Override
-    public void visit(SuperSection superSection) {
-
-        visitChildren(superSection);
-    }
-
-
-    @Override
-    public void visit(Section section) {
-
-        visitChildren(section);
-    }
-
-
-    @Override
-    public void visit(Headline headline) {
-
-        visitChildren(headline);
-    }
-
-
-    @Override
-    public void visit(Paragraph paragraph) {
-
-        visitChildren(paragraph);
-    }
-
-
-    @Override
-    public void visit(Footer footer) {
-
-        visitChildren(footer);
-    }
-
-
-    @Override
-    public void visit(Header header) {
-
-        visitChildren(header);
-    }
-
-
-    @Override
-    public void visit(Image image) {
-
-        visitChildren(image);
-    }
-
-
-    @Override
-    public void visit(Table table) {
-
-        visitChildren(table);
-    }
-
-
-    @Override
-    public void visit(TableCell tableCell) {
-
-        visitChildren(tableCell);
-    }
-
-
-    protected void visitChildren(SemanticNode semanticNode) {
-
-        semanticNode.streamChildren()
-                .forEach(node -> node.accept(this));
-
-    }
-
-}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/DocumentTree.java

@@ -2,13 +2,12 @@ package com.knecon.fforesight.service.layoutparser.processor.model.graph;
 
 import static java.lang.String.format;
 
-import java.util.ArrayList;
 import java.util.Collections;
 import java.util.LinkedList;
 import java.util.List;
 import java.util.stream.Stream;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeTypeProto.NodeType;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Document;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.GenericSemanticNode;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.SemanticNode;
@@ -40,10 +39,7 @@ public class DocumentTree {
 
     public TextBlock buildTextBlock() {
 
-        return allEntriesInOrder().map(Entry::getNode)
-                .filter(SemanticNode::isLeaf)
-                .map(SemanticNode::getLeafTextBlock)
-                .collect(new TextBlockCollector());
+        return allEntriesInOrder().map(Entry::getNode).filter(SemanticNode::isLeaf).map(SemanticNode::getLeafTextBlock).collect(new TextBlockCollector());
     }
 
 
@@ -117,16 +113,13 @@ public class DocumentTree {
 
     public Stream<SemanticNode> childNodes(List<Integer> treeId) {
 
-        return getEntryById(treeId).children.stream()
-                .map(Entry::getNode);
+        return getEntryById(treeId).children.stream().map(Entry::getNode);
     }
 
 
     public Stream<SemanticNode> childNodesOfType(List<Integer> treeId, NodeType nodeType) {
 
-        return getEntryById(treeId).children.stream()
-                .filter(entry -> entry.node.getType().equals(nodeType))
-                .map(Entry::getNode);
+        return getEntryById(treeId).children.stream().filter(entry -> entry.node.getType().equals(nodeType)).map(Entry::getNode);
     }
 
 
@@ -163,32 +156,26 @@ public class DocumentTree {
 
     public Stream<Entry> allEntriesInOrder() {
 
-        return Stream.of(root)
-                .flatMap(DocumentTree::flatten);
+        return Stream.of(root).flatMap(DocumentTree::flatten);
     }
 
 
     public Stream<Entry> allSubEntriesInOrder(List<Integer> parentId) {
 
-        return getEntryById(parentId).children.stream()
-                .flatMap(DocumentTree::flatten);
+        return getEntryById(parentId).children.stream().flatMap(DocumentTree::flatten);
     }
 
 
     @Override
     public String toString() {
 
-        return String.join("\n",
-                           allEntriesInOrder().map(Entry::toString)
-                                   .toList());
+        return String.join("\n", allEntriesInOrder().map(Entry::toString).toList());
     }
 
 
     private static Stream<Entry> flatten(Entry entry) {
 
-        return Stream.concat(Stream.of(entry),
-                             entry.children.stream()
-                                     .flatMap(DocumentTree::flatten));
+        return Stream.concat(Stream.of(entry), entry.children.stream().flatMap(DocumentTree::flatten));
     }
 
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/NodeVisitor.java

@@ -1,45 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.processor.model.graph;
-
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Document;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Footer;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Header;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Headline;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Image;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Paragraph;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Section;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.SuperSection;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Table;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.TableCell;
-
-public interface NodeVisitor {
-
-    void visit(Document document);
-
-
-    void visit(SuperSection superSection);
-
-
-    void visit(Section section);
-
-
-    void visit(Headline headline);
-
-
-    void visit(Paragraph paragraph);
-
-
-    void visit(Footer footer);
-
-
-    void visit(Header header);
-
-
-    void visit(Image image);
-
-
-    void visit(Table table);
-
-
-    void visit(TableCell tableCell);
-
-}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/TextRange.java

@@ -79,12 +79,6 @@ public class TextRange implements Comparable<TextRange> {
 
     public boolean contains(int index) {
 
-        return start <= index && index <= end;
-    }
-
-
-    public boolean containsExclusive(int index) {
-
         return start <= index && index < end;
     }
 
@@ -97,13 +91,8 @@ public class TextRange implements Comparable<TextRange> {
 
     public List<TextRange> split(List<Integer> splitIndices) {
 
-        if (splitIndices.stream()
-                .anyMatch(idx -> !this.contains(idx))) {
-            throw new IndexOutOfBoundsException(format("%s splitting indices are out of range for %s",
-                                                       splitIndices.stream()
-                                                               .filter(idx -> !this.contains(idx))
-                                                               .toList(),
-                                                       this));
+        if (splitIndices.stream().anyMatch(idx -> !this.contains(idx))) {
+            throw new IndexOutOfBoundsException(format("%s splitting indices are out of range for %s", splitIndices.stream().filter(idx -> !this.contains(idx)).toList(), this));
         }
         List<TextRange> splitBoundaries = new LinkedList<>();
         int previousIndex = start;
@@ -116,29 +105,19 @@ public class TextRange implements Comparable<TextRange> {
             splitBoundaries.add(new TextRange(previousIndex, splitIndex));
             previousIndex = splitIndex;
         }
-        if (previousIndex != end) {
-            splitBoundaries.add(new TextRange(previousIndex, end));
-        }
+        splitBoundaries.add(new TextRange(previousIndex, end));
         return splitBoundaries;
     }
 
-
     public IntStream intStream() {
 
         return IntStream.range(start, end);
     }
 
-
     public static TextRange merge(Collection<TextRange> boundaries) {
 
-        int minStart = boundaries.stream()
-                .mapToInt(TextRange::start)
-                .min()
-                .orElseThrow(IllegalArgumentException::new);
-        int maxEnd = boundaries.stream()
-                .mapToInt(TextRange::end)
-                .max()
-                .orElseThrow(IllegalArgumentException::new);
+        int minStart = boundaries.stream().mapToInt(TextRange::start).min().orElseThrow(IllegalArgumentException::new);
+        int maxEnd = boundaries.stream().mapToInt(TextRange::end).max().orElseThrow(IllegalArgumentException::new);
         return new TextRange(minStart, maxEnd);
     }
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/entity/RedactionEntity.java

@@ -28,7 +28,7 @@ import lombok.experimental.FieldDefaults;
 @AllArgsConstructor
 @FieldDefaults(level = AccessLevel.PRIVATE)
 @EqualsAndHashCode(onlyExplicitlyIncluded = true)
-public class TextEntity {
+public class RedactionEntity {
 
     // initial values
     @EqualsAndHashCode.Include
@@ -47,7 +47,7 @@ public class TextEntity {
     boolean dictionaryEntry;
     boolean dossierDictionaryEntry;
     Set<Engine> engines;
-    Set<TextEntity> references;
+    Set<RedactionEntity> references;
     @Builder.Default
     Deque<Integer> matchedRules = new LinkedList<>();
     String redactionReason;
@@ -66,9 +66,9 @@ public class TextEntity {
     SemanticNode deepestFullyContainingNode;
 
 
-    public static TextEntity initialEntityNode(TextRange textRange, String type, EntityType entityType) {
+    public static RedactionEntity initialEntityNode(TextRange textRange, String type, EntityType entityType) {
 
-        return TextEntity.builder().type(type).entityType(entityType).textRange(textRange).engines(new HashSet<>()).references(new HashSet<>()).build();
+        return RedactionEntity.builder().type(type).entityType(entityType).textRange(textRange).engines(new HashSet<>()).references(new HashSet<>()).build();
     }
 
 
@@ -155,21 +155,21 @@ public class TextEntity {
     }
 
 
-    public boolean containedBy(TextEntity textEntity) {
+    public boolean containedBy(RedactionEntity redactionEntity) {
 
-        return this.textRange.containedBy(textEntity.getTextRange());
+        return this.textRange.containedBy(redactionEntity.getTextRange());
     }
 
 
-    public boolean contains(TextEntity textEntity) {
+    public boolean contains(RedactionEntity redactionEntity) {
 
-        return this.textRange.contains(textEntity.getTextRange());
+        return this.textRange.contains(redactionEntity.getTextRange());
     }
 
 
-    public boolean intersects(TextEntity textEntity) {
+    public boolean intersects(RedactionEntity redactionEntity) {
 
-        return this.textRange.intersects(textEntity.getTextRange());
+        return this.textRange.intersects(redactionEntity.getTextRange());
     }
 
 
@@ -185,13 +185,13 @@ public class TextEntity {
     }
 
 
-    public void addReference(TextEntity reference) {
+    public void addReference(RedactionEntity reference) {
 
         references.add(reference);
     }
 
 
-    public void addReferences(List<TextEntity> references) {
+    public void addReferences(List<RedactionEntity> references) {
 
         this.references.addAll(references);
     }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/AbstractSemanticNode.java

@@ -6,9 +6,9 @@ import java.util.List;
 import java.util.Map;
 import java.util.Set;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.LayoutEngineProto.LayoutEngine;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.LayoutEngine;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.DocumentTree;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.entity.TextEntity;
+import com.knecon.fforesight.service.layoutparser.processor.model.graph.entity.RedactionEntity;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.TextBlock;
 
 import lombok.AccessLevel;
@@ -39,7 +39,7 @@ public abstract class AbstractSemanticNode implements GenericSemanticNode {
 
     @Builder.Default
     @EqualsAndHashCode.Exclude
-    Set<TextEntity> entities = new HashSet<>();
+    Set<RedactionEntity> entities = new HashSet<>();
 
     @EqualsAndHashCode.Exclude
     Map<Page, Rectangle2D> bBoxCache;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/Document.java

@@ -9,11 +9,10 @@ import java.util.Set;
 import java.util.stream.Collectors;
 import java.util.stream.Stream;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeTypeProto.NodeType;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.DocumentTree;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.NodeVisitor;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.TextBlock;
-import com.knecon.fforesight.service.layoutparser.processor.visualization.LayoutDebugLayer;
+import com.knecon.fforesight.service.layoutparser.processor.visualization.LayoutparsingVisualizations;
 
 import lombok.AccessLevel;
 import lombok.AllArgsConstructor;
@@ -34,7 +33,7 @@ public class Document extends AbstractSemanticNode {
     Set<Page> pages;
     Integer numberOfPages;
 
-    LayoutDebugLayer layoutDebugLayer;
+    LayoutparsingVisualizations visualizations;
 
 
     @Override
@@ -102,7 +101,8 @@ public class Document extends AbstractSemanticNode {
     public Headline getHeadline() {
 
         return streamAllSubNodesOfType(NodeType.HEADLINE).map(node -> (Headline) node)
-                .findFirst().orElse(Headline.builder().build());
+                .findFirst()
+                .orElse(Headline.builder().build());
     }
 
 
@@ -163,11 +163,4 @@ public class Document extends AbstractSemanticNode {
         return bBox;
     }
 
-
-    @Override
-    public void accept(NodeVisitor visitor) {
-
-        visitor.visit(this);
-    }
-
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/DuplicatedParagraph.java

@@ -25,4 +25,11 @@ public class DuplicatedParagraph extends Paragraph {
 
     }
 
+
+    @Override
+    public String toString() {
+
+        return super.toString();
+    }
+
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/Footer.java

@@ -1,7 +1,6 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeTypeProto.NodeType;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.NodeVisitor;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.TextBlock;
 
 import lombok.AccessLevel;
@@ -35,12 +34,6 @@ public class Footer extends AbstractSemanticNode {
     }
 
 
-    @Override
-    public void accept(NodeVisitor visitor) {
-        visitor.visit(this);
-    }
-
-
     @Override
     public TextBlock getTextBlock() {
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/Header.java

@@ -1,7 +1,6 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeTypeProto.NodeType;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.NodeVisitor;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.TextBlock;
 
 import lombok.AccessLevel;
@@ -28,13 +27,6 @@ public class Header extends AbstractSemanticNode {
     }
 
 
-    @Override
-    public void accept(NodeVisitor visitor) {
-
-        visitor.visit(this);
-    }
-
-
     @Override
     public NodeType getType() {
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/Headline.java

@@ -1,7 +1,6 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeTypeProto.NodeType;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.NodeVisitor;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.TextBlock;
 
 import lombok.AccessLevel;
@@ -35,13 +34,6 @@ public class Headline extends AbstractSemanticNode {
     }
 
 
-    @Override
-    public void accept(NodeVisitor visitor) {
-
-        visitor.visit(this);
-    }
-
-
     @Override
     public TextBlock getTextBlock() {
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/Image.java

@@ -6,8 +6,7 @@ import java.util.HashMap;
 import java.util.Map;
 import java.util.Set;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeTypeProto.NodeType;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.NodeVisitor;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.TextBlock;
 
 import lombok.AccessLevel;
@@ -29,8 +28,6 @@ public class Image extends AbstractSemanticNode {
 
     String id;
 
-    String representationHash;
-
     ImageType imageType;
     boolean transparent;
     Rectangle2D position;
@@ -87,29 +84,10 @@ public class Image extends AbstractSemanticNode {
     }
 
 
-    @Override
-    public void accept(NodeVisitor visitor) {
-
-        visitor.visit(this);
-    }
-
-
     @Override
     public boolean isLeaf() {
 
         return true;
     }
 
-
-    public double getArea() {
-
-        return position.getWidth() * position.getHeight();
-    }
-
-
-    public boolean isFullPageImage() {
-
-        return imageType.equals(ImageType.OCR) || getArea() >= 0.5 * page.getArea();
-    }
-
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/ImageType.java

@@ -6,6 +6,7 @@ public enum ImageType {
     LOGO,
     FORMULA,
     SIGNATURE,
+
     SIGNATURE_VISUAL,
     OTHER,
     OCR,

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/Page.java

@@ -1,15 +1,12 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes;
 
-import java.util.Comparator;
 import java.util.HashSet;
 import java.util.LinkedList;
 import java.util.List;
 import java.util.Set;
-import java.util.stream.Stream;
 
 import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationPage;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.entity.TextEntity;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.AtomicTextBlock;
+import com.knecon.fforesight.service.layoutparser.processor.model.graph.entity.RedactionEntity;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.TextBlock;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.TextBlockCollector;
 
@@ -32,8 +29,9 @@ public class Page {
     Integer height;
     Integer width;
     Integer rotation;
+
     @EqualsAndHashCode.Exclude
-    List<AtomicTextBlock> textBlocksOnPage;
+    List<SemanticNode> mainBody;
     @EqualsAndHashCode.Exclude
     Header header;
     @EqualsAndHashCode.Exclude
@@ -41,7 +39,7 @@ public class Page {
 
     @Builder.Default
     @EqualsAndHashCode.Exclude
-    Set<TextEntity> entities = new HashSet<>();
+    Set<RedactionEntity> entities = new HashSet<>();
 
     @Builder.Default
     @EqualsAndHashCode.Exclude
@@ -55,41 +53,14 @@ public class Page {
                 .width((int) classificationPage.getPageWidth())
                 .number(classificationPage.getPageNumber())
                 .rotation(classificationPage.getRotation())
-                .textBlocksOnPage(new LinkedList<>())
+                .mainBody(new LinkedList<>())
                 .build();
     }
 
 
-    /**
-     * Constructs and returns a {@link TextBlock} representing the concatenated text of all leaf semantic nodes in the main body.
-     *
-     * @return The main body text block.
-     */
     public TextBlock getMainBodyTextBlock() {
 
-        return textBlocksOnPage.stream()
-                .filter(atb -> !atb.isEmpty())
-                .collect(new TextBlockCollector());
-    }
-
-
-    public List<SemanticNode> getMainBody() {
-
-        return textBlocksOnPage.stream()
-                .map(AtomicTextBlock::getParent)
-                .map(this::getHighestParentOnPage)
-                .distinct()
-                .toList();
-    }
-
-
-    private SemanticNode getHighestParentOnPage(SemanticNode node) {
-
-        SemanticNode currentNode = node;
-        while (currentNode.hasParent() && currentNode.getParent().onlyOnPage(this)) {
-            currentNode = currentNode.getParent();
-        }
-        return currentNode;
+        return mainBody.stream().filter(SemanticNode::isLeaf).map(SemanticNode::getLeafTextBlock).collect(new TextBlockCollector());
     }
 
 
@@ -113,10 +84,4 @@ public class Page {
         return o instanceof Page && o.hashCode() == this.hashCode();
     }
 
-
-    public double getArea() {
-
-        return height * width;
-    }
-
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/Paragraph.java

@@ -1,7 +1,6 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeTypeProto.NodeType;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.NodeVisitor;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.TextBlock;
 
 import lombok.AccessLevel;
@@ -35,13 +34,6 @@ public class Paragraph extends AbstractSemanticNode {
     }
 
 
-    @Override
-    public void accept(NodeVisitor visitor) {
-
-        visitor.visit(this);
-    }
-
-
     @Override
     public TextBlock getTextBlock() {
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/Section.java

@@ -1,8 +1,6 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes;
 
-
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeTypeProto.NodeType;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.NodeVisitor;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
 
 import lombok.AccessLevel;
 import lombok.AllArgsConstructor;
@@ -28,20 +26,15 @@ public class Section extends AbstractSemanticNode {
     public Headline getHeadline() {
 
         return streamChildrenOfType(NodeType.HEADLINE).map(node -> (Headline) node)
-                .findFirst().orElseGet(() -> getParent().getHeadline());
-    }
-
-
-    @Override
-    public void accept(NodeVisitor visitor) {
-
-        visitor.visit(this);
+                .findFirst()
+                .orElseGet(() -> getParent().getHeadline());
     }
 
 
     public boolean hasTables() {
 
-        return streamAllSubNodesOfType(NodeType.TABLE).findAny().isPresent();
+        return streamAllSubNodesOfType(NodeType.TABLE).findAny()
+                .isPresent();
     }
 
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/SemanticNode.java

@@ -12,18 +12,16 @@ import java.util.Set;
 import java.util.stream.Collectors;
 import java.util.stream.Stream;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.LayoutEngineProto.LayoutEngine;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeTypeProto.NodeType;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.DocumentTree;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.LayoutEngine;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.TextRange;
+import com.knecon.fforesight.service.layoutparser.processor.model.graph.DocumentTree;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.entity.EntityType;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.entity.TextEntity;
+import com.knecon.fforesight.service.layoutparser.processor.model.graph.entity.RedactionEntity;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.AtomicTextBlock;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.TextBlock;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.TextBlockCollector;
-import com.knecon.fforesight.service.layoutparser.processor.utils.BBoxMergingUtility;
 import com.knecon.fforesight.service.layoutparser.processor.utils.RectangleTransformations;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.NodeVisitor;
 
 public interface SemanticNode {
 
@@ -56,7 +54,7 @@ public interface SemanticNode {
      *
      * @return Set of all Entities associated with this Node
      */
-    Set<TextEntity> getEntities();
+    Set<RedactionEntity> getEntities();
 
 
     /**
@@ -223,7 +221,7 @@ public interface SemanticNode {
      * @param type string representing the type of entities to return
      * @return List of RedactionEntities of any the type
      */
-    default List<TextEntity> getEntitiesOfType(String type) {
+    default List<RedactionEntity> getEntitiesOfType(String type) {
 
         return getEntities().stream()
                 .filter(redactionEntity -> redactionEntity.getType().equals(type))
@@ -237,7 +235,7 @@ public interface SemanticNode {
      * @param types A list of strings representing the types of entities to return
      * @return List of RedactionEntities of any provided type
      */
-    default List<TextEntity> getEntitiesOfType(List<String> types) {
+    default List<RedactionEntity> getEntitiesOfType(List<String> types) {
 
         return getEntities().stream()
                 .filter(redactionEntity -> redactionEntity.isAnyType(types))
@@ -255,7 +253,8 @@ public interface SemanticNode {
 
         TextBlock textBlock = getTextBlock();
         if (!textBlock.getAtomicTextBlocks().isEmpty()) {
-            return getTextBlock().getAtomicTextBlocks().get(0).getNumberOnPage();
+            return getTextBlock().getAtomicTextBlocks()
+                    .get(0).getNumberOnPage();
         } else {
             return -1;
         }
@@ -340,19 +339,19 @@ public interface SemanticNode {
      * This function is used during insertion of EntityNodes into the graph, it checks if the boundary of the RedactionEntity intersects or even contains the RedactionEntity.
      * It sets the fields accordingly and recursively calls this function on all its children.
      *
-     * @param textEntity RedactionEntity, which is being inserted into the graph
+     * @param redactionEntity RedactionEntity, which is being inserted into the graph
      */
-    default void addThisToEntityIfIntersects(TextEntity textEntity) {
+    default void addThisToEntityIfIntersects(RedactionEntity redactionEntity) {
 
         TextBlock textBlock = getTextBlock();
-        if (textBlock.getTextRange().intersects(textEntity.getTextRange())) {
-            if (textBlock.containsBoundary(textEntity.getTextRange())) {
-                textEntity.setDeepestFullyContainingNode(this);
+        if (textBlock.getTextRange().intersects(redactionEntity.getTextRange())) {
+            if (textBlock.containsBoundary(redactionEntity.getTextRange())) {
+                redactionEntity.setDeepestFullyContainingNode(this);
             }
 
-            textEntity.addIntersectingNode(this);
-            streamChildren().filter(semanticNode -> semanticNode.getBoundary().intersects(textEntity.getTextRange()))
-                    .forEach(node -> node.addThisToEntityIfIntersects(textEntity));
+            redactionEntity.addIntersectingNode(this);
+            streamChildren().filter(semanticNode -> semanticNode.getBoundary().intersects(redactionEntity.getTextRange()))
+                    .forEach(node -> node.addThisToEntityIfIntersects(redactionEntity));
         }
     }
 
@@ -473,19 +472,34 @@ public interface SemanticNode {
      */
     private Map<Page, Rectangle2D> getBBoxFromChildren() {
 
-        List<Map<Page, Rectangle2D>> childrenBBoxes = streamChildren().filter(child -> !isFullPageImage(child))
+        Map<Page, Rectangle2D> bBoxPerPage = new HashMap<>();
+
+        List<Map<Page, Rectangle2D>> childrenBBoxes = streamChildren() //
+                .filter(SemanticNode::isNotOcrImage)
                 .map(SemanticNode::getBBox)
                 .toList();
-        return BBoxMergingUtility.mergeBBoxes(childrenBBoxes);
+
+        Set<Page> pages = childrenBBoxes.stream()
+                .flatMap(map -> map.keySet()
+                        .stream())
+                .collect(Collectors.toSet());
+        for (Page page : pages) {
+            Rectangle2D bBoxOnPage = childrenBBoxes.stream()
+                    .filter(childBboxPerPage -> childBboxPerPage.containsKey(page))
+                    .map(childBboxPerPage -> childBboxPerPage.get(page))
+                    .collect(RectangleTransformations.collectBBox());
+            bBoxPerPage.put(page, bBoxOnPage);
+        }
+        return bBoxPerPage;
     }
 
 
-    private static boolean isFullPageImage(SemanticNode child) {
+    private static boolean isNotOcrImage(SemanticNode node) {
 
-        if (!child.getType().equals(NodeType.IMAGE)) {
-            return false;
+        if (!node.getType().equals(NodeType.IMAGE)) {
+            return true;
         }
-        return ((Image) child).isFullPageImage();
+        return false;
     }
 
 
@@ -502,20 +516,4 @@ public interface SemanticNode {
         return bBoxPerPage;
     }
 
-
-    void accept(NodeVisitor visitor);
-
-
-    /**
-     * Checks wether this SemanticNode appears on a single page only, and if that page is the provided one.
-     *
-     * @param page the page to check
-     * @return true, when SemanticNode is on a single page only and the page is the provided page. Otherwise, false.
-     */
-    default boolean onlyOnPage(Page page) {
-
-        Set<Page> pages = getPages();
-        return pages.size() == 1 && pages.contains(page);
-    }
-
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/SuperSection.java

@@ -1,7 +1,6 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeTypeProto.NodeType;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.NodeVisitor;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
 
 import lombok.AccessLevel;
 import lombok.AllArgsConstructor;
@@ -27,14 +26,8 @@ public class SuperSection extends AbstractSemanticNode {
     public Headline getHeadline() {
 
         return streamChildrenOfType(NodeType.HEADLINE).map(node -> (Headline) node)
-                .findFirst().orElseGet(() -> getParent().getHeadline());
-    }
-
-
-    @Override
-    public void accept(NodeVisitor visitor) {
-
-        visitor.visit(this);
+                .findFirst()
+                .orElseGet(() -> getParent().getHeadline());
     }
 
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/Table.java

@@ -12,11 +12,10 @@ import java.util.Set;
 import java.util.stream.IntStream;
 import java.util.stream.Stream;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.LayoutEngineProto.LayoutEngine;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeTypeProto.NodeType;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.LayoutEngine;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.DocumentTree;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.NodeVisitor;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.entity.TextEntity;
+import com.knecon.fforesight.service.layoutparser.processor.model.graph.entity.RedactionEntity;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.TextBlock;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.TextBlockCollector;
 
@@ -44,7 +43,7 @@ public class Table implements SemanticNode {
 
     @Builder.Default
     @EqualsAndHashCode.Exclude
-    Set<TextEntity> entities = new HashSet<>();
+    Set<RedactionEntity> entities = new HashSet<>();
 
     @EqualsAndHashCode.Exclude
     Map<Page, Rectangle2D> bBoxCache;
@@ -55,7 +54,7 @@ public class Table implements SemanticNode {
      * @param strings Strings to check whether a row contains them
      * @return Stream of all entities in this table, that appear in a row, which contains any of the provided strings
      */
-    public Stream<TextEntity> streamEntitiesWhereRowContainsStringsIgnoreCase(List<String> strings) {
+    public Stream<RedactionEntity> streamEntitiesWhereRowContainsStringsIgnoreCase(List<String> strings) {
 
         return IntStream.range(0, numberOfRows).boxed()
                 .filter(row -> rowContainsStringsIgnoreCase(row, strings))
@@ -89,7 +88,7 @@ public class Table implements SemanticNode {
      * @param value  the string which the table cell should contain
      * @return a stream of all entities, which appear in a row where at least one cell has the provided header and the provided value.
      */
-    public Stream<TextEntity> streamEntitiesWhereRowHasHeaderAndValue(String header, String value) {
+    public Stream<RedactionEntity> streamEntitiesWhereRowHasHeaderAndValue(String header, String value) {
 
         List<Integer> vertebrateStudyCols = streamHeaders().filter(headerNode -> headerNode.containsString(header))
                 .map(TableCell::getCol)
@@ -108,7 +107,7 @@ public class Table implements SemanticNode {
      * @param values the strings which the table cell should contain
      * @return a stream of all entities, which appear in a row where at least one cell has the provided header and any provided value.
      */
-    public Stream<TextEntity> streamEntitiesWhereRowHasHeaderAndAnyValue(String header, List<String> values) {
+    public Stream<RedactionEntity> streamEntitiesWhereRowHasHeaderAndAnyValue(String header, List<String> values) {
 
         List<Integer> colsWithHeader = streamHeaders().filter(headerNode -> headerNode.containsString(header))
                 .map(TableCell::getCol)
@@ -126,7 +125,7 @@ public class Table implements SemanticNode {
      * @param types type strings to check whether a row contains an entity like them
      * @return Stream of all entities in this table, that appear in a row, which contains at least one entity with any of the provided types.
      */
-    public Stream<TextEntity> streamEntitiesWhereRowContainsEntitiesOfType(List<String> types) {
+    public Stream<RedactionEntity> streamEntitiesWhereRowContainsEntitiesOfType(List<String> types) {
 
         List<Integer> rowsWithEntityOfType = IntStream.range(0, numberOfRows).boxed()
                 .filter(rowNumber -> streamEntityTypesInRow(rowNumber).anyMatch(existingType -> types.stream()
@@ -146,7 +145,7 @@ public class Table implements SemanticNode {
      * @param types type strings to check whether a row doesn't contain an entity like it
      * @return Stream of all entities in this table, that appear in a row, which does not contain any entity with any of the provided types.
      */
-    public Stream<TextEntity> streamEntitiesWhereRowContainsNoEntitiesOfType(List<String> types) {
+    public Stream<RedactionEntity> streamEntitiesWhereRowContainsNoEntitiesOfType(List<String> types) {
 
         List<Integer> rowsWithNoEntityOfType = IntStream.range(0, numberOfRows).boxed()
                 .filter(rowNumber -> streamEntityTypesInRow(rowNumber).noneMatch(existingType -> types.stream()
@@ -164,7 +163,7 @@ public class Table implements SemanticNode {
 
         return streamRow(rowNumber).map(TableCell::getEntities)
                 .flatMap(Collection::stream)
-                .map(TextEntity::getType)
+                .map(RedactionEntity::getType)
                 .distinct();
     }
 
@@ -305,12 +304,12 @@ public class Table implements SemanticNode {
      * Finds all entities of the provided type, which appear in the same row that the provided entity appears in.
      *
      * @param type            the type of entities to search for
-     * @param textEntity the entity, which appears in the row to search
+     * @param redactionEntity the entity, which appears in the row to search
      * @return List of all entities of the provided type, which appear in the same row that the provided entity appears in.
      */
-    public List<TextEntity> getEntitiesOfTypeInSameRow(String type, TextEntity textEntity) {
+    public List<RedactionEntity> getEntitiesOfTypeInSameRow(String type, RedactionEntity redactionEntity) {
 
-        return textEntity.getIntersectingNodes()
+        return redactionEntity.getIntersectingNodes()
                 .stream()
                 .filter(node -> node instanceof TableCell)
                 .map(node -> (TableCell) node)
@@ -354,10 +353,4 @@ public class Table implements SemanticNode {
         return bBoxCache;
     }
 
-
-    @Override
-    public void accept(NodeVisitor visitor) {
-        visitor.visit(this);
-    }
-
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/TableCell.java

@@ -4,8 +4,7 @@ import java.awt.geom.Rectangle2D;
 import java.util.HashMap;
 import java.util.Map;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeTypeProto.NodeType;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.NodeVisitor;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.TextBlock;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.TextBlockCollector;
 
@@ -43,13 +42,6 @@ public class TableCell extends AbstractSemanticNode {
     }
 
 
-    @Override
-    public void accept(NodeVisitor visitor) {
-
-        visitor.visit(this);
-    }
-
-
     @Override
     public NodeType getType() {
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/textblock/AtomicTextBlock.java

@@ -1,6 +1,5 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock;
 
-import static com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentTextDataProto.DocumentTextData;
 import static java.lang.String.format;
 
 import java.awt.geom.Rectangle2D;
@@ -14,8 +13,8 @@ import java.util.Map;
 import java.util.Set;
 import java.util.stream.Collectors;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentPositionDataProto.DocumentPositionData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentPositionDataProto.DocumentPositionData.Position;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentPositionData;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentTextData;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.TextRange;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Page;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.SemanticNode;
@@ -75,7 +74,7 @@ public class AtomicTextBlock implements TextBlock {
         }
 
         Set<Integer> lbInBoundary = lineBreaks.stream()
-                .map(i -> i + this.textRange.start())
+                .map(i -> i + stringTextRange.start())
                 .filter(stringTextRange::contains)
                 .collect(Collectors.toSet());
         if (stringTextRange.end() == getTextRange().end()) {
@@ -154,9 +153,11 @@ public class AtomicTextBlock implements TextBlock {
                 .page(page)
                 .textRange(new TextRange(documentTextData.getStart(), documentTextData.getEnd()))
                 .searchText(documentTextData.getSearchText())
-                .lineBreaks(documentTextData.getLineBreaksList())
-                .stringIdxToPositionIdx(documentPositionData.getStringIdxToPositionIdxList())
-                .positions(toRectangle2DList(documentPositionData.getPositionsList()))
+                .lineBreaks(Arrays.stream(documentTextData.getLineBreaks()).boxed()
+                                    .toList())
+                .stringIdxToPositionIdx(Arrays.stream(documentPositionData.getStringIdxToPositionIdx()).boxed()
+                                                .toList())
+                .positions(toRectangle2DList(documentPositionData.getPositions()))
                 .parent(parent)
                 .build();
     }
@@ -170,23 +171,12 @@ public class AtomicTextBlock implements TextBlock {
     }
 
 
-    private static List<Rectangle2D> toRectangle2DList(List<Position> positions) {
-
-        return positions.stream()
-                .map(pos -> (Rectangle2D) new Rectangle2D.Float(pos.getValue(0), pos.getValue(1), pos.getValue(2), pos.getValue(3)))
-                .toList();
-    }
-
-
     public CharSequence getLine(int lineNumber) {
 
         if (lineNumber >= numberOfLines() || lineNumber < 0) {
             throw new IndexOutOfBoundsException(format("line %d out of range for AtomicTextBlock with %d lines", lineNumber, numberOfLines()));
         }
         if (lineNumber == 0) {
-            if (lineBreaks.isEmpty()) {
-                return searchText;
-            }
             return subSequence(textRange.start(), lineBreaks.get(0) + textRange.start());
         } else if (lineNumber == numberOfLines() - 1) {
             return subSequence(lineBreaks.get(lineBreaks.size() - 1) + textRange.start(), textRange.end());
@@ -264,7 +254,7 @@ public class AtomicTextBlock implements TextBlock {
     }
 
 
-    protected List<Integer> getAllLineBreaksInBoundary(TextRange textRange) {
+    private List<Integer> getAllLineBreaksInBoundary(TextRange textRange) {
 
         return getLineBreaks().stream()
                 .map(linebreak -> linebreak + this.textRange.start())

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/textblock/ConcatenatedTextBlock.java

@@ -69,7 +69,7 @@ public class ConcatenatedTextBlock implements TextBlock {
     private AtomicTextBlock getAtomicTextBlockByStringIndex(int stringIdx) {
 
         return atomicTextBlocks.stream()
-                .filter(textBlock -> textBlock.getTextRange().containsExclusive(stringIdx))
+                .filter(textBlock -> textBlock.getTextRange().contains(stringIdx))
                 .findAny()
                 .orElseThrow(IndexOutOfBoundsException::new);
     }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/textblock/TextBlock.java

@@ -115,7 +115,7 @@ public interface TextBlock extends CharSequence {
 
     default boolean containsIndex(int stringIndex) {
 
-        return getTextRange().containsExclusive(stringIndex);
+        return getTextRange().contains(stringIndex);
     }
 
 

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

@@ -10,8 +10,8 @@ import lombok.NonNull;
 import lombok.RequiredArgsConstructor;
 
 @Data
-@AllArgsConstructor
 @RequiredArgsConstructor
+@AllArgsConstructor
 public class ClassifiedImage {
 
     @NonNull
@@ -20,18 +20,9 @@ public class ClassifiedImage {
     private ImageType imageType;
     private boolean sourceByAi;
     private boolean isAppendedToSection;
+    @NonNull
     private boolean hasTransparency;
+    @NonNull
     private int page;
-    private String representation;
-
-
-    public ClassifiedImage(@NonNull Rectangle2D position, @NonNull ImageType imageType, boolean hasTransparency, int page, String representation) {
-
-        this.position = position;
-        this.imageType = imageType;
-        this.hasTransparency = hasTransparency;
-        this.page = page;
-        this.representation = representation;
-    }
 
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/outline/OutlineExtractorService.java

@@ -1,6 +1,5 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.outline;
 
-import java.awt.geom.AffineTransform;
 import java.awt.geom.Point2D;
 import java.io.IOException;
 import java.util.ArrayList;
@@ -27,9 +26,6 @@ import org.apache.pdfbox.pdmodel.interactive.documentnavigation.outline.PDDocume
 import org.apache.pdfbox.pdmodel.interactive.documentnavigation.outline.PDOutlineItem;
 import org.springframework.stereotype.Service;
 
-import com.knecon.fforesight.service.layoutparser.processor.utils.CoordinateTransforms;
-import com.knecon.fforesight.service.layoutparser.processor.utils.PageInformation;
-
 import lombok.SneakyThrows;
 import lombok.extern.slf4j.Slf4j;
 
@@ -86,20 +82,11 @@ public class OutlineExtractorService {
 
         String title = item.getTitle();
 
-        PDPage page;
-        try {
-            // Can throw: "Error: can't convert to Destination COSArray" for some OCR'd PDFs
-            page = item.findDestinationPage(document);
-            if (page == null) {
-                return Optional.empty();
-            }
-        } catch (IOException e) {
-            log.info(String.format("Error occurred during position resolution for outline item with title %s: " + e, title));
+        PDPage page = item.findDestinationPage(document);
+        if (page == null) {
             return Optional.empty();
         }
-
-        int pageNumber = document.getPages().indexOf(page) + 1;
-        AffineTransform userSpaceToPageCoords = CoordinateTransforms.calculateInitialUserSpaceCoordsToPageCoords(PageInformation.fromPDPage(pageNumber, page));
+        int pageNumber = document.getPages().indexOf(page);
 
         Optional<Point2D> outlinePosition = Optional.empty();
 
@@ -128,15 +115,8 @@ public class OutlineExtractorService {
             log.info(String.format("Error occurred during position resolution for outline item on page %s with title %s: " + e, pageNumber, title));
         }
 
-        return Optional.of(new OutlineObjectTreeNode(new OutlineObject(title,
-                                                                       pageNumber,
-                                                                       transformPointToPageCoords(outlinePosition, userSpaceToPageCoords), depth)));
-    }
+        return Optional.of(new OutlineObjectTreeNode(new OutlineObject(title, pageNumber, outlinePosition.orElse(new Point2D.Float(0, 0)), depth)));
 
-
-    private static Point2D transformPointToPageCoords(Optional<Point2D> outlinePosition, AffineTransform userSpaceToPageCoords) {
-
-        return outlinePosition.map(point -> userSpaceToPageCoords.transform(point, null)).orElse(null);
     }
 
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/outline/OutlineObject.java

@@ -1,34 +1,27 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.outline;
 
 import java.awt.geom.Point2D;
-import java.util.Optional;
 
-import com.knecon.fforesight.service.layoutparser.processor.docstrum.model.BoundingBox;
-
-import lombok.Getter;
-import lombok.Setter;
+import lombok.AllArgsConstructor;
+import lombok.Data;
+import lombok.RequiredArgsConstructor;
 
+@Data
+@RequiredArgsConstructor
+@AllArgsConstructor
 public class OutlineObject {
 
-    @Getter
     private final String title;
-    @Getter
     private final int pageNumber;
-    @Getter
+    private Point2D point;
     private final int treeDepth;
 
-    private Point2D point; // java coordinates, (0, 0) is always top left
-
-    @Getter
-    @Setter
     private boolean found;
 
 
     public OutlineObject(String title, int pageNumber, Point2D point2D, int depth) {
 
-        this.title = title;
-        this.pageNumber = pageNumber;
-        this.treeDepth = depth;
+        this(title, pageNumber, depth);
         this.point = point2D;
     }
 
@@ -39,39 +32,4 @@ public class OutlineObject {
         return "OutlineObject{" + "title='" + title + '\'' + '}';
     }
 
-
-    public Optional<Point2D> getPoint() {
-
-        return Optional.ofNullable(point);
-    }
-
-
-    public boolean isAbove(BoundingBox boundingBox) {
-
-        if (point == null) {
-            return true;
-        }
-        return point.getY() <= boundingBox.getMaxY();
-    }
-
-
-    public double distance(BoundingBox boundingBox) {
-
-        if (point == null) {
-            return 0;
-        }
-        if (boundingBox.getBBox().contains(point)) {
-            return 0;
-        }
-        double deltaX = Math.min(Math.abs(boundingBox.getMinX() - point.getX()), Math.abs(boundingBox.getMaxX() - point.getX()));
-        double deltaY = Math.min(Math.abs(boundingBox.getMinY() - point.getY()), Math.abs(boundingBox.getMaxY() - point.getY()));
-        return Math.sqrt(deltaX * deltaX + deltaY * deltaY);
-    }
-
-
-    public void resetPoint() {
-
-        this.point = null;
-    }
-
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/outline/OutlineObjectTree.java

@@ -39,28 +39,4 @@ public class OutlineObjectTree {
         }
     }
 
-    @Override
-    public String toString() {
-
-        StringBuilder sb = new StringBuilder();
-        sb.append("OutlineObjectTree(\n");
-        for (OutlineObjectTreeNode node : rootNodes) {
-            buildString(node, sb, 1);
-        }
-        sb.append(")");
-        return sb.toString();
-    }
-
-    private void buildString(OutlineObjectTreeNode node, StringBuilder sb, int depth) {
-
-        for (int i = 0; i < depth; i++) {
-            sb.append("  ");
-        }
-        sb.append(node.getOutlineObject().getTitle()).append("\n");
-
-        for (OutlineObjectTreeNode child : node.getChildren()) {
-            buildString(child, sb, depth + 1);
-        }
-    }
-
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/outline/OutlineValidationService.java

@@ -10,20 +10,15 @@ import java.util.TreeSet;
 
 import org.springframework.stereotype.Service;
 
-import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationDocument;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
 
-import io.micrometer.observation.annotation.Observed;
 import lombok.extern.slf4j.Slf4j;
 
 @Service
 @Slf4j
 public class OutlineValidationService {
 
-    @Observed(name = "OutlineValidationService", contextualName = "create-toc")
-    public TableOfContents createToC(ClassificationDocument classificationDocument) {
-
-        List<TextPageBlock> headlines = extractHeadlines(classificationDocument);
+    public TableOfContents createToC(List<TextPageBlock> headlines) {
 
         List<TableOfContentItem> mainSections = new ArrayList<>();
         Map<Integer, TableOfContentItem> lastItemsPerDepth = new HashMap<>();
@@ -63,16 +58,4 @@ public class OutlineValidationService {
         return new TableOfContents(mainSections);
     }
 
-
-    private static List<TextPageBlock> extractHeadlines(ClassificationDocument classificationDocument) {
-
-        return classificationDocument.getPages()
-                .stream()
-                .flatMap(classificationPage -> classificationPage.getTextBlocks()
-                        .stream()
-                        .filter(tb -> tb instanceof TextPageBlock && tb.getClassification() != null && tb.getClassification().isHeadline())
-                        .map(tb -> (TextPageBlock) tb))
-                .toList();
-    }
-
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/outline/TOCEnrichmentService.java

@@ -1,7 +1,6 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.outline;
 
 import java.util.ArrayList;
-import java.util.Collection;
 import java.util.Collections;
 import java.util.Iterator;
 import java.util.List;
@@ -168,7 +167,7 @@ public class TOCEnrichmentService {
             }
         }
 
-        if (!startBlocks.isEmpty() || !startImages.isEmpty()) {
+        if (!startBlocks.isEmpty()) {
             TableOfContentItem unassigned = new TableOfContentItem(null);
             unassigned.setSectionBlocks(startBlocks);
             unassigned.setImages(startImages);
@@ -186,8 +185,12 @@ public class TOCEnrichmentService {
             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)
+            if (previousTableNonHeaderRow.isEmpty()
+                && previousTable.getRowCount() == 1
+                && previousTable.getRows()
+                           .get(0).size() == tableNonHeaderRow.size()) {
+                previousTableNonHeaderRow = previousTable.getRows()
+                        .get(0)
                         .stream()
                         .map(cell -> {
                             Cell fakeCell = Cell.copy(cell);
@@ -198,7 +201,8 @@ public class TOCEnrichmentService {
             }
             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<Cell> 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++) {
@@ -221,15 +225,18 @@ public class TOCEnrichmentService {
 
         return table.getRows()
                 .stream()
-                .flatMap(Collection::stream)
-                .allMatch(cell -> cell.getHeaderCells().isEmpty());
+                .flatMap(row -> row.stream()
+                        .filter(cell -> !cell.getHeaderCells().isEmpty()))
+                .findAny().isEmpty();
+
     }
 
 
     private List<Cell> getRowWithNonHeaderCells(TablePageBlock table) {
 
         for (int i = table.getRowCount() - 1; i >= 0; i--) { // Non header rows are most likely at bottom of table
-            List<Cell> row = table.getRows().get(i);
+            List<Cell> row = table.getRows()
+                    .get(i);
             if (row.size() == 1) {
                 continue;
             }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/outline/TableOfContentItem.java

@@ -5,7 +5,7 @@ import java.util.List;
 import java.util.stream.Collectors;
 
 import com.knecon.fforesight.service.layoutparser.processor.model.AbstractPageBlock;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.GenericSemanticNode;
+import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.AbstractSemanticNode;
 import com.knecon.fforesight.service.layoutparser.processor.model.image.ClassifiedImage;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
 
@@ -24,7 +24,7 @@ public class TableOfContentItem {
     private List<AbstractPageBlock> sectionBlocks = new ArrayList<>();
     private List<ClassifiedImage> images = new ArrayList<>();
 
-    private GenericSemanticNode section;
+    private AbstractSemanticNode section;
 
 
     public TableOfContentItem(TextPageBlock headline) {
@@ -45,7 +45,8 @@ public class TableOfContentItem {
         if (parent != null) {
             int index = parent.getChildren().indexOf(this);
             if (index > 0) {
-                return parent.getChildren().get(index - 1);
+                return parent.getChildren()
+                        .get(index - 1);
             }
         }
         return null;
@@ -57,7 +58,8 @@ public class TableOfContentItem {
         if (parent != null) {
             int index = parent.getChildren().indexOf(this);
             if (index >= 0 && index < parent.getChildren().size() - 1) {
-                return parent.getChildren().get(index + 1);
+                return parent.getChildren()
+                        .get(index + 1);
             }
         }
         return null;
@@ -91,19 +93,17 @@ public class TableOfContentItem {
         return false;
     }
 
-
     public List<AbstractPageBlock> getNonEmptySectionBlocks() {
 
-        return sectionBlocks.stream()
-                .filter(pageBlock -> !pageBlock.isEmpty())
-                .collect(Collectors.toList());
+        return sectionBlocks.stream().filter(pageBlock -> !pageBlock.isEmpty()).collect(Collectors.toList());
     }
 
-
     @Override
     public String toString() {
 
         return "OutlineObjectTreeNode{" + "textPageBlock=" + headline + '}';
     }
 
+
+
 }

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

@@ -9,7 +9,7 @@ import java.util.List;
 
 import com.knecon.fforesight.service.layoutparser.processor.docstrum.model.BoundingBox;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
-import com.knecon.fforesight.service.layoutparser.processor.model.text.Word;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPositionSequence;
 import com.knecon.fforesight.service.layoutparser.processor.utils.TextNormalizationUtilities;
 
 import lombok.Data;
@@ -35,14 +35,14 @@ public class Cell extends BoundingBox {
 
     public Cell(Point2D topLeft, Point2D bottomRight) {
 
-        this.bBoxPdf = new Rectangle2D.Double(topLeft.getX(), topLeft.getY(), (bottomRight.getX() - topLeft.getX()), (bottomRight.getY() - topLeft.getY()));
-        this.bBox = bBoxPdf;
+        this.bBoxInitialUserSpace = new Rectangle2D.Double(topLeft.getX(), topLeft.getY(), (bottomRight.getX() - topLeft.getX()), (bottomRight.getY() - topLeft.getY()));
+        this.bBox = bBoxInitialUserSpace;
     }
 
 
     public Cell(Rectangle2D bBoxInitialUserSpace, AffineTransform initialUserSpaceToJava) {
 
-        this.bBoxPdf = bBoxInitialUserSpace;
+        this.bBoxInitialUserSpace = bBoxInitialUserSpace;
         this.bBox = initialUserSpaceToJava.createTransformedShape(bBoxInitialUserSpace).getBounds2D();
     }
 
@@ -50,7 +50,7 @@ public class Cell extends BoundingBox {
     public static Cell copy(Cell cell) {
 
         Cell copy = new Cell();
-        copy.bBoxPdf = cell.bBoxPdf;
+        copy.bBoxInitialUserSpace = cell.bBoxInitialUserSpace;
         copy.bBox = cell.bBox;
         return copy;
     }
@@ -68,12 +68,12 @@ public class Cell extends BoundingBox {
         StringBuilder sb = new StringBuilder();
 
         Iterator<TextPageBlock> itty = textBlocks.iterator();
-        Word previous = null;
+        TextPositionSequence previous = null;
         while (itty.hasNext()) {
 
             TextPageBlock textBlock = itty.next();
 
-            for (Word word : textBlock.getWords()) {
+            for (TextPositionSequence word : textBlock.getSequences()) {
                 if (previous != null) {
                     if (Math.abs(previous.getMaxYDirAdj() - word.getMaxYDirAdj()) > word.getTextHeight()) {
                         sb.append('\n');
@@ -87,7 +87,7 @@ public class Cell extends BoundingBox {
 
         }
 
-        return TextNormalizationUtilities.cleanString(sb.toString());
+        return TextNormalizationUtilities.removeHyphenLineBreaks(sb.toString()).replaceAll("\n", " ").replaceAll(" {2}", " ");
     }
 
 

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

@@ -70,7 +70,7 @@ public class CleanRulings {
 
     public boolean lineBetween(BoundingBox a, BoundingBox b) {
 
-        return lineBetween(a.getBBoxPdf(), b.getBBoxPdf());
+        return lineBetween(a.getBBoxInitialUserSpace(), b.getBBoxInitialUserSpace());
     }
 
 

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

@@ -263,8 +263,8 @@ public class TablePageBlock extends AbstractPageBlock {
 
                         cells.stream()
                                 .map(originalCell -> new CellWithIntersection(originalCell,
-                                                                              RectangleTransformations.calculateIntersectedArea(cellFromGridStructure.getBBoxPdf(),
-                                                                                                                                originalCell.getBBoxPdf())))
+                                                                              RectangleTransformations.calculateIntersectedArea(cellFromGridStructure.getBBoxInitialUserSpace(),
+                                                                                                                                originalCell.getBBoxInitialUserSpace())))
                                 .filter(cellWithIntersection -> cellWithIntersection.intersectedArea > 0)
                                 .filter(cellWithIntersection -> cellWithIntersection.originalCell.getArea() > cellWithIntersection.intersectedArea * CELL_AREA_CONTAINED_THRESHOLD)
                                 .max(Comparator.comparing(CellWithIntersection::intersectedArea))

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

@@ -1,8 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.processor.model.text;
-
-import com.knecon.fforesight.service.layoutparser.processor.model.AbstractPageBlock;
-import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationPage;
-
-public record AbstractBlockOnPage(AbstractPageBlock block, ClassificationPage page) {
-
-}

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

@@ -1,9 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.processor.model.text;
-
-
-public enum FontStyle {
-    REGULAR,
-    BOLD,
-    ITALIC,
-    BOLD_ITALIC;
-}

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

@@ -1,21 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.processor.model.text;
-
-import com.knecon.fforesight.service.layoutparser.processor.model.FloatFrequencyCounter;
-
-import lombok.AccessLevel;
-import lombok.Getter;
-import lombok.NoArgsConstructor;
-import lombok.experimental.FieldDefaults;
-
-@Getter
-@NoArgsConstructor
-@FieldDefaults(makeFinal = true, level = AccessLevel.PRIVATE)
-public class FrequencyCounters {
-
-    FloatFrequencyCounter lineHeightFrequencyCounter = new FloatFrequencyCounter();
-    FloatFrequencyCounter fontSizeFrequencyCounter = new FloatFrequencyCounter();
-    FloatFrequencyCounter spaceFrequencyCounter = new FloatFrequencyCounter();
-    StringFrequencyCounter fontFrequencyCounter = new StringFrequencyCounter();
-    StringFrequencyCounter styleFrequencyCounter = new StringFrequencyCounter();
-
-}

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

@@ -1,107 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.processor.model.text;
-
-import java.util.List;
-import java.util.Optional;
-import java.util.regex.Matcher;
-import java.util.regex.Pattern;
-
-import lombok.AccessLevel;
-import lombok.AllArgsConstructor;
-import lombok.Getter;
-import lombok.experimental.FieldDefaults;
-
-@AllArgsConstructor
-@FieldDefaults(makeFinal = true, level = AccessLevel.PRIVATE)
-public class ListIdentifier {
-
-    public static final Pattern NUMBER_WITH_DOT = Pattern.compile("^\\s*([1-9]{1,4})\\.\\s+");
-    public static final Pattern NUMBER_IN_PARENTHESES = Pattern.compile("^\\s*\\(([1-9]{1,4})\\)\\s+");
-
-    enum Format {
-        NUMBER_WITH_DOT,
-        NUMBER_IN_PARENTHESES
-    }
-
-    Format format;
-    @Getter
-    Word word;
-    @Getter
-    int page;
-    int representation;
-
-
-    public static Optional<ListIdentifier> parse(TextPageBlock textPageBlock, int page) {
-
-        return parse(textPageBlock.getWords().subList(0, Math.min(5, textPageBlock.getWords().size())), page);
-    }
-
-
-    public static Optional<ListIdentifier> parse(List<Word> sequences, int page) {
-
-        StringBuilder sb = new StringBuilder();
-        for (Word sequence : sequences) {
-            sb.append(sequence.toString());
-            sb.append(" ");
-        }
-        sb.replace(sb.length() - 1, sb.length(), "");
-        String text = sb.toString();
-
-        Matcher numberMatcher = NUMBER_WITH_DOT.matcher(text);
-
-        if (numberMatcher.find()) {
-            Optional<Integer> representation = parseInteger(numberMatcher.group(1));
-            if (representation.isPresent()) {
-                return Optional.of(new ListIdentifier(Format.NUMBER_WITH_DOT, sequences.get(0), page, representation.get()));
-            }
-        }
-
-        Matcher parenthesisMatcher = NUMBER_IN_PARENTHESES.matcher(text);
-        if (parenthesisMatcher.find()) {
-            Optional<Integer> representation = parseInteger(parenthesisMatcher.group(1));
-            if (representation.isPresent()) {
-                return Optional.of(new ListIdentifier(Format.NUMBER_IN_PARENTHESES, sequences.get(0), page, representation.get()));
-            }
-        }
-        return Optional.empty();
-    }
-
-
-    private static Optional<Integer> parseInteger(String text) {
-
-        try {
-            return Optional.of(Integer.parseInt(text));
-        } catch (NumberFormatException e) {
-            return Optional.empty();
-        }
-    }
-
-
-    public static boolean isInOrder(List<ListIdentifier> listIdentifiers) {
-
-        if (listIdentifiers.size() <= 1) {
-            return true;
-        }
-
-        for (int i = 1; i < listIdentifiers.size(); i++) {
-            ListIdentifier current = listIdentifiers.get(i);
-            ListIdentifier previous = listIdentifiers.get(i - 1);
-            if (current.format != previous.format) {
-                return false;
-            }
-            if (current.representation <= previous.representation) {
-                return false;
-            }
-            if (!current.word.intersectsXDirAdj(previous.word, 2)) {
-                return false;
-            }
-            if (current.page == previous.page && !current.word.isBelowDirAdj(previous.word)) {
-                return false;
-            }
-            if (current.page < previous.page) {
-                return false;
-            }
-        }
-        return true;
-    }
-
-}