Merge branch 'hotfix' into 'main'

Hotfix

See merge request fforesight/layout-parser!243

.gitlab-ci.yml

@@ -21,5 +21,6 @@ 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

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

@@ -7,4 +7,5 @@ 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/redaction/DocumentData.java

@@ -2,6 +2,11 @@ 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;
@@ -13,16 +18,26 @@ 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.")
-    DocumentPage[] documentPages;
+    AllDocumentPages documentPages;
     @Schema(description = "Contains information about the document's text.")
-    DocumentTextData[] documentTextData;
+    AllDocumentTextData documentTextData;
     @Schema(description = "Contains information about the document's text positions.")
-    DocumentPositionData[] documentPositions;
+    AllDocumentPositionData documentPositions;
     @Schema(description = "Contains information about the document's semantic structure.")
-    DocumentStructure documentStructure;
+    DocumentStructureWrapper documentStructureWrapper;
+
+
+    public DocumentStructure getDocumentStructure() {
+
+        return documentStructureWrapper.getDocumentStructure();
+    }
 
 }

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

@@ -10,6 +10,7 @@ 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

@@ -10,6 +10,7 @@ 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

@@ -16,6 +16,7 @@ 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/DocumentStructureProto.java

@@ -0,0 +1,799 @@
+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

@@ -0,0 +1,126 @@
+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

@@ -10,6 +10,7 @@ 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

@@ -1,5 +1,6 @@
 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

@@ -0,0 +1,193 @@
+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

@@ -3,6 +3,7 @@ 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

@@ -0,0 +1,274 @@
+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/resources/DocumentPage.proto

@@ -0,0 +1,21 @@
+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

@@ -0,0 +1,25 @@
+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

@@ -0,0 +1,8 @@
+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

@@ -0,0 +1,29 @@
+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

@@ -0,0 +1,27 @@
+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

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

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

@@ -0,0 +1,14 @@
+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

@@ -12,12 +12,14 @@ dependencies {
     implementation(project(":layoutparser-service-internal-api"))
     implementation(project(":viewer-doc-processor"))
 
-    implementation("com.iqser.red.service:persistence-service-shared-api-v1:2.144.0") {
+    implementation("com.iqser.red.service:persistence-service-shared-api-v1:2.564.0-RED9010.0") {
         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")
-    implementation("com.iqser.red.commons:storage-commons:2.45.0")
+    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("org.apache.pdfbox:pdfbox:${pdfBoxVersion}")
     implementation("org.apache.pdfbox:pdfbox-tools:${pdfBoxVersion}")
@@ -34,4 +36,5 @@ dependencies {
     implementation("com.pdftron:PDFNet:10.11.0")
     implementation("org.apache.commons:commons-text:1.12.0")
 
+    implementation("com.google.protobuf:protobuf-java-util:4.27.1")
 }

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

@@ -2,11 +2,13 @@ package com.knecon.fforesight.service.layoutparser.processor;
 
 import static java.lang.String.format;
 
+import java.awt.geom.AffineTransform;
 import java.awt.geom.Rectangle2D;
 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;
@@ -20,10 +22,14 @@ import org.apache.pdfbox.pdmodel.common.PDRectangle;
 import org.apache.pdfbox.pdmodel.documentinterchange.markedcontent.PDMarkedContent;
 import org.springframework.stereotype.Service;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeTypeProto.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.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.classification.TableOfContentsClassificationService;
 import com.knecon.fforesight.service.layoutparser.processor.services.mapper.MarkdownMapper;
 import com.knecon.fforesight.service.layoutparser.processor.model.AbstractPageBlock;
 import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationDocument;
@@ -91,10 +97,7 @@ 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 +107,12 @@ 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 {
@@ -273,6 +276,9 @@ public class LayoutParsingPipeline {
             stripper.setPdpage(pdPage);
             stripper.getText(originDocument);
             List<TextPositionSequence> words = stripper.getTextPositionSequences();
+
+//            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);
@@ -366,33 +372,9 @@ public class LayoutParsingPipeline {
 
         originDocument.close();
 
-        log.info("Calculating BodyTextFrame for {}", identifier);
-        bodyTextFrameService.setBodyTextFrames(classificationDocument, layoutParsingType);
-        for (ClassificationPage page : classificationDocument.getPages()) {
-            classificationDocument.getLayoutDebugLayer().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);
-        }
+        classificationService.classify(classificationDocument, layoutParsingType, identifier);
 
-        if (layoutParsingType.equals(LayoutParsingType.DOCUMINE_OLD)) {
-            for (ClassificationPage page : classificationDocument.getPages()) {
-                docuMineBlockificationService.mergeblocks(page, page.getCleanRulings().withoutTextRulings(), 0, 10);
-            }
-        }
-
-        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);
+        TableOfContents tableOfContents = outlineValidationService.createToC(classificationDocument);
         classificationDocument.setTableOfContents(tableOfContents);
 
         log.info("Building Sections for {}", identifier);
@@ -406,6 +388,32 @@ public class LayoutParsingPipeline {
     }
 
 
+    private static void rotateDirAdjExactly(List<TextPositionSequence> words, PDPage pdPage) {
+
+        for (TextDirection dir : TextDirection.values()) {
+
+            double averageRotation = words.stream()
+                    .map(TextPositionSequence::getTextPositions)
+                    .flatMap(Collection::stream)
+                    .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 (TextPositionSequence word : words) {
+                if (!dir.equals(word.getDir())) {
+                    continue;
+                }
+                word.transform(rotateInstance);
+            }
+        }
+    }
+
+
     private void addNumberOfPagesToTrace(int numberOfPages, long size) {
 
         if (observationRegistry.getCurrentObservation() != null) {

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

@@ -11,7 +11,9 @@ 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;
@@ -39,6 +41,8 @@ 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 {
@@ -100,13 +104,35 @@ public class LayoutParsingStorageService {
     }
 
 
+    @SneakyThrows
     @Observed(name = "LayoutParsingStorageService", contextualName = "store-document-data")
     public void storeDocumentData(LayoutParsingRequest layoutParsingRequest, DocumentData documentData) {
 
-        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());
+        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();
     }
 
 

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

@@ -225,33 +225,31 @@ public abstract class BoundingBox {
 
     public double horizontalDistance(BoundingBox other) {
 
-        Rectangle2D left;
-        Rectangle2D right;
-        if (this.leftOf(other)) {
-            left = this.getBBox();
-            right = other.getBBox();
-        } else {
-            left = other.getBBox();
-            right = this.getBBox();
-        }
+        double rect1Right = getMaxX();
+        double rect1Left = getMinX();
+        double rect2Right = other.getMaxX();
+        double rect2Left = other.getMinX();
 
-        return Math.max(0, right.getMinX() - left.getMaxX());
+        if (rect1Left > rect2Right || rect2Left > rect1Right) {
+            return Math.max(rect2Left - rect1Right, rect1Left - rect2Right);
+        } else {
+            return 0;
+        }
     }
 
 
     public double verticalDistance(BoundingBox other) {
 
-        Rectangle2D bottom;
-        Rectangle2D top;
-        if (this.isAbove(other)) {
-            top = this.getBBox();
-            bottom = other.getBBox();
-        } else {
-            bottom = this.getBBox();
-            top = other.getBBox();
-        }
+        double rect1Top = getMaxY();
+        double rect1Bottom = getMinY();
+        double rect2Top = other.getMaxY();
+        double rect2Bottom = other.getMinY();
 
-        return Math.max(0, bottom.getMinY() - top.getMaxY());
+        if (rect1Bottom > rect2Top || rect2Bottom > rect1Top) {
+            return Math.max(rect2Bottom - rect1Top, rect1Bottom - rect2Top);
+        } else {
+            return 0;
+        }
     }
 
 

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

@@ -99,4 +99,82 @@ public abstract class TextBoundingBox extends BoundingBox {
         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

@@ -28,4 +28,10 @@ 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/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.LayoutEngine;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.LayoutEngineProto.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/FloatFrequencyCounter.java

@@ -44,14 +44,14 @@ public class FloatFrequencyCounter {
 
     public Double getMostPopular() {
 
-        if (changed) {
+        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;
                 }
             }
-            mostPopularCache = mostPopular != null ? mostPopular.getKey() : null;
+            mostPopularCache = mostPopular != null ? mostPopular.getKey() : 0;
             changed = false;
         }
 

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

@@ -9,12 +9,13 @@ 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/graph/DocumentTree.java

@@ -8,7 +8,7 @@ import java.util.LinkedList;
 import java.util.List;
 import java.util.stream.Stream;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeTypeProto.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;

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

@@ -98,10 +98,10 @@ public class TextRange implements Comparable<TextRange> {
     public List<TextRange> split(List<Integer> splitIndices) {
 
         if (splitIndices.stream()
-                .anyMatch(idx -> !this.containsExclusive(idx))) {
+                .anyMatch(idx -> !this.contains(idx))) {
             throw new IndexOutOfBoundsException(format("%s splitting indices are out of range for %s",
                                                        splitIndices.stream()
-                                                               .filter(idx -> !this.containsExclusive(idx))
+                                                               .filter(idx -> !this.contains(idx))
                                                                .toList(),
                                                        this));
         }
@@ -116,7 +116,9 @@ public class TextRange implements Comparable<TextRange> {
             splitBoundaries.add(new TextRange(previousIndex, splitIndex));
             previousIndex = splitIndex;
         }
-        splitBoundaries.add(new TextRange(previousIndex, end));
+        if (previousIndex != end) {
+            splitBoundaries.add(new TextRange(previousIndex, end));
+        }
         return splitBoundaries;
     }
 

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

@@ -6,7 +6,7 @@ import java.util.List;
 import java.util.Map;
 import java.util.Set;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.LayoutEngine;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.LayoutEngineProto.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.textblock.TextBlock;

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

@@ -9,7 +9,7 @@ import java.util.Set;
 import java.util.stream.Collectors;
 import java.util.stream.Stream;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
+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.processor.model.graph.NodeVisitor;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.TextBlock;

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

@@ -1,6 +1,6 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
+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.processor.model.graph.textblock.TextBlock;
 

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

@@ -1,6 +1,6 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
+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.processor.model.graph.textblock.TextBlock;
 

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

@@ -1,6 +1,6 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
+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.processor.model.graph.textblock.TextBlock;
 

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

@@ -6,7 +6,7 @@ import java.util.HashMap;
 import java.util.Map;
 import java.util.Set;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
+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.processor.model.graph.textblock.TextBlock;
 

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

@@ -1,6 +1,6 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
+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.processor.model.graph.textblock.TextBlock;
 

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

@@ -1,6 +1,7 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
+
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeTypeProto.NodeType;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.NodeVisitor;
 
 import lombok.AccessLevel;

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

@@ -12,8 +12,8 @@ import java.util.Set;
 import java.util.stream.Collectors;
 import java.util.stream.Stream;
 
-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.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.processor.model.graph.TextRange;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.entity.EntityType;

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

@@ -1,6 +1,6 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeTypeProto.NodeType;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.NodeVisitor;
 
 import lombok.AccessLevel;

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

@@ -12,8 +12,8 @@ import java.util.Set;
 import java.util.stream.IntStream;
 import java.util.stream.Stream;
 
-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.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.processor.model.graph.NodeVisitor;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.entity.TextEntity;

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

@@ -4,7 +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.NodeType;
+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.processor.model.graph.textblock.TextBlock;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.TextBlockCollector;

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

@@ -1,5 +1,6 @@
 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;
@@ -13,8 +14,8 @@ import java.util.Map;
 import java.util.Set;
 import java.util.stream.Collectors;
 
-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.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.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;
@@ -153,11 +154,9 @@ public class AtomicTextBlock implements TextBlock {
                 .page(page)
                 .textRange(new TextRange(documentTextData.getStart(), documentTextData.getEnd()))
                 .searchText(documentTextData.getSearchText())
-                .lineBreaks(Arrays.stream(documentTextData.getLineBreaks()).boxed()
-                                    .toList())
-                .stringIdxToPositionIdx(Arrays.stream(documentPositionData.getStringIdxToPositionIdx()).boxed()
-                                                .toList())
-                .positions(toRectangle2DList(documentPositionData.getPositions()))
+                .lineBreaks(documentTextData.getLineBreaksList())
+                .stringIdxToPositionIdx(documentPositionData.getStringIdxToPositionIdxList())
+                .positions(toRectangle2DList(documentPositionData.getPositionsList()))
                 .parent(parent)
                 .build();
     }
@@ -171,6 +170,14 @@ 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) {

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

@@ -10,6 +10,7 @@ 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;
@@ -20,7 +21,9 @@ import lombok.extern.slf4j.Slf4j;
 public class OutlineValidationService {
 
     @Observed(name = "OutlineValidationService", contextualName = "create-toc")
-    public TableOfContents createToC(List<TextPageBlock> headlines) {
+    public TableOfContents createToC(ClassificationDocument classificationDocument) {
+
+        List<TextPageBlock> headlines = extractHeadlines(classificationDocument);
 
         List<TableOfContentItem> mainSections = new ArrayList<>();
         Map<Integer, TableOfContentItem> lastItemsPerDepth = new HashMap<>();
@@ -60,4 +63,16 @@ 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/text/AbstractBlockOnPage.java

@@ -0,0 +1,8 @@
+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/FrequencyCounters.java

@@ -0,0 +1,21 @@
+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

@@ -0,0 +1,90 @@
+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]+)\\.\\s+");
+    public static final Pattern NUMBER_IN_PARENTHESES = Pattern.compile("^\\s*\\(([1-9]+)\\)\\s+");
+
+    enum Format {
+        NUMBER_WITH_DOT, NUMBER_IN_PARENTHESES
+    }
+
+    Format format;
+    @Getter
+    TextPositionSequence word;
+    @Getter
+    int page;
+    int representation;
+
+
+    public static Optional<ListIdentifier> parse(TextPageBlock textPageBlock, int page) {
+
+        return parse(textPageBlock.getSequences().subList(0, Math.min(5, textPageBlock.getSequences().size())), page);
+    }
+
+
+    public static Optional<ListIdentifier> parse(List<TextPositionSequence> sequences, int page) {
+
+        StringBuilder sb = new StringBuilder();
+        for (TextPositionSequence 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()) {
+            return Optional.of(new ListIdentifier(Format.NUMBER_WITH_DOT, sequences.get(0), page, Integer.parseInt(numberMatcher.group(1))));
+        }
+
+        Matcher parenthesisMatcher = NUMBER_IN_PARENTHESES.matcher(text);
+        if (parenthesisMatcher.find()) {
+            return Optional.of(new ListIdentifier(Format.NUMBER_IN_PARENTHESES, sequences.get(0), page, Integer.parseInt(parenthesisMatcher.group(1))));
+        }
+        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;
+    }
+
+}

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

@@ -7,6 +7,8 @@ import org.apache.pdfbox.text.TextPosition;
 
 import com.knecon.fforesight.service.layoutparser.processor.docstrum.model.TextBoundingBox;
 import com.knecon.fforesight.service.layoutparser.processor.docstrum.utils.FastAtan2;
+import com.knecon.fforesight.service.layoutparser.processor.utils.CoordinateTransforms;
+import com.knecon.fforesight.service.layoutparser.processor.utils.PageInformation;
 
 import lombok.AccessLevel;
 import lombok.AllArgsConstructor;

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

@@ -9,10 +9,14 @@ public class StringFrequencyCounter {
 
     @Getter
     private final Map<String, Integer> countPerValue = new HashMap<>();
+    boolean changed;
+    String mostPopularCache;
 
 
     public void add(String value) {
 
+        changed = true;
+
         if (!countPerValue.containsKey(value)) {
             countPerValue.put(value, 1);
         } else {
@@ -23,6 +27,8 @@ public class StringFrequencyCounter {
 
     public void addAll(Map<String, Integer> otherCounter) {
 
+        changed = true;
+
         for (Map.Entry<String, Integer> entry : otherCounter.entrySet()) {
             if (countPerValue.containsKey(entry.getKey())) {
                 countPerValue.put(entry.getKey(), countPerValue.get(entry.getKey()) + entry.getValue());
@@ -35,13 +41,18 @@ public class StringFrequencyCounter {
 
     public String getMostPopular() {
 
-        Map.Entry<String, Integer> mostPopular = null;
-        for (Map.Entry<String, Integer> entry : countPerValue.entrySet()) {
-            if (mostPopular == null || entry.getValue() > mostPopular.getValue()) {
-                mostPopular = entry;
+        if (changed || mostPopularCache == null) {
+            Map.Entry<String, Integer> mostPopular = null;
+            for (Map.Entry<String, Integer> entry : countPerValue.entrySet()) {
+                if (mostPopular == null || entry.getValue() >= mostPopular.getValue()) {
+                    mostPopular = entry;
+                }
             }
+            mostPopularCache = mostPopular != null ? mostPopular.getKey() : null;
+            changed = false;
         }
-        return mostPopular != null ? mostPopular.getKey() : null;
+
+        return mostPopularCache;
     }
 
 }

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

@@ -0,0 +1,7 @@
+package com.knecon.fforesight.service.layoutparser.processor.model.text;
+
+import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationPage;
+
+public record TextBlockOnPage(TextPageBlock textBlock, ClassificationPage page) {
+
+}

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

@@ -7,7 +7,6 @@ import java.util.List;
 
 import com.fasterxml.jackson.annotation.JsonIgnore;
 import com.knecon.fforesight.service.layoutparser.processor.model.AbstractPageBlock;
-import com.knecon.fforesight.service.layoutparser.processor.model.FloatFrequencyCounter;
 import com.knecon.fforesight.service.layoutparser.processor.model.PageBlockType;
 import com.knecon.fforesight.service.layoutparser.processor.utils.RectangleTransformations;
 import com.knecon.fforesight.service.layoutparser.processor.utils.TextNormalizationUtilities;
@@ -27,23 +26,13 @@ public class TextPageBlock extends AbstractPageBlock {
 
     @Builder.Default
     private List<TextPositionSequence> sequences = new ArrayList<>();
+    @Builder.Default
+    private FrequencyCounters frequencyCounters = new FrequencyCounters();
 
     private Rectangle2D bBoxDirAdj;
 
-    private String mostPopularWordFont;
-
-    private String mostPopularWordStyle;
-
-    private double mostPopularWordFontSize;
-
-    private double mostPopularWordHeight;
-
-    private double mostPopularWordSpaceWidth;
-
     private boolean underlined;
 
-    private double highestFontSize;
-
     private PageBlockType classification;
 
     private boolean toDuplicate;
@@ -55,8 +44,10 @@ public class TextPageBlock extends AbstractPageBlock {
     public TextPageBlock(List<TextPositionSequence> sequences) {
 
         this.sequences = new ArrayList<>(sequences);
+        this.frequencyCounters = new FrequencyCounters();
+
         if (!sequences.isEmpty()) {
-            calculateFrequencyCounters();
+            addToFrequencyCounters(sequences);
         }
         calculateBBox();
     }
@@ -118,32 +109,18 @@ public class TextPageBlock extends AbstractPageBlock {
     }
 
 
-    private void calculateFrequencyCounters() {
-
-        FloatFrequencyCounter lineHeightFrequencyCounter = new FloatFrequencyCounter();
-        FloatFrequencyCounter fontSizeFrequencyCounter = new FloatFrequencyCounter();
-        FloatFrequencyCounter spaceFrequencyCounter = new FloatFrequencyCounter();
-        StringFrequencyCounter fontFrequencyCounter = new StringFrequencyCounter();
-        StringFrequencyCounter styleFrequencyCounter = new StringFrequencyCounter();
+    private void addToFrequencyCounters(List<TextPositionSequence> sequences) {
 
         for (TextPositionSequence wordBlock : sequences) {
 
-            lineHeightFrequencyCounter.add(wordBlock.getTextHeight());
-            fontSizeFrequencyCounter.add(wordBlock.getFontSize());
-            spaceFrequencyCounter.add(wordBlock.getSpaceWidth());
-            fontFrequencyCounter.add(wordBlock.getFont());
-            styleFrequencyCounter.add(wordBlock.getFontStyle());
-
+            frequencyCounters.getLineHeightFrequencyCounter().add(wordBlock.getTextHeight());
+            frequencyCounters.getFontSizeFrequencyCounter().add(wordBlock.getFontSize());
+            frequencyCounters.getSpaceFrequencyCounter().add(wordBlock.getSpaceWidth());
+            frequencyCounters.getFontFrequencyCounter().add(wordBlock.getFont());
+            frequencyCounters.getStyleFrequencyCounter().add(wordBlock.getFontStyle());
         }
 
-        setMostPopularWordFont(fontFrequencyCounter.getMostPopular());
-        setMostPopularWordStyle(styleFrequencyCounter.getMostPopular());
-        setMostPopularWordFontSize(fontSizeFrequencyCounter.getMostPopular());
-        setMostPopularWordHeight(lineHeightFrequencyCounter.getMostPopular());
-        setMostPopularWordSpaceWidth(spaceFrequencyCounter.getMostPopular());
-        setHighestFontSize(fontSizeFrequencyCounter.getHighest());
-
-        setUnderlined(sequences.stream()
+        setUnderlined(this.sequences.stream()
                               .allMatch(TextPositionSequence::isUnderline));
     }
 
@@ -152,7 +129,7 @@ public class TextPageBlock extends AbstractPageBlock {
 
         TextPageBlock union = this.copy();
         union.add(r);
-        calculateFrequencyCounters();
+        addToFrequencyCounters(List.of(r));
         calculateBBox();
         return union;
     }
@@ -162,7 +139,7 @@ public class TextPageBlock extends AbstractPageBlock {
 
         TextPageBlock union = this.copy();
         union.addAll(r.getSequences());
-        calculateFrequencyCounters();
+        addToFrequencyCounters(r.getSequences());
         calculateBBox();
         return union;
     }
@@ -172,7 +149,7 @@ public class TextPageBlock extends AbstractPageBlock {
 
         changed = true;
         sequences.addAll(textPageBlock.getSequences());
-        calculateFrequencyCounters();
+        addToFrequencyCounters(textPageBlock.getSequences());
         calculateBBox();
     }
 
@@ -181,7 +158,7 @@ public class TextPageBlock extends AbstractPageBlock {
 
         changed = true;
         sequences.add(textPositionSequence);
-        calculateFrequencyCounters();
+        addToFrequencyCounters(List.of(textPositionSequence));
         calculateBBox();
     }
 
@@ -190,7 +167,7 @@ public class TextPageBlock extends AbstractPageBlock {
 
         changed = true;
         sequences.addAll(textPositionSequences);
-        calculateFrequencyCounters();
+        addToFrequencyCounters(textPositionSequences);
         calculateBBox();
     }
 
@@ -253,6 +230,43 @@ public class TextPageBlock extends AbstractPageBlock {
     }
 
 
+    public String getMostPopularWordFont() {
+
+        return frequencyCounters.getFontFrequencyCounter().getMostPopular();
+    }
+
+
+    public String getMostPopularWordStyle() {
+
+        return frequencyCounters.getStyleFrequencyCounter().getMostPopular();
+    }
+
+
+    public double getMostPopularWordFontSize() {
+
+        return frequencyCounters.getFontSizeFrequencyCounter().getMostPopular();
+    }
+
+
+    public double getMostPopularWordHeight() {
+
+        return frequencyCounters.getLineHeightFrequencyCounter().getMostPopular();
+    }
+
+
+    public double getMostPopularWordSpaceWidth() {
+
+        return frequencyCounters.getSpaceFrequencyCounter().getMostPopular();
+    }
+
+
+    public double getHighestFontSize() {
+
+        Double highest = frequencyCounters.getFontSizeFrequencyCounter().getHighest();
+        return highest == null ? 0 : highest;
+    }
+
+
     @Override
     public boolean isEmpty() {
 

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

@@ -2,10 +2,13 @@ package com.knecon.fforesight.service.layoutparser.processor.model.text;
 
 import static com.knecon.fforesight.service.layoutparser.processor.model.text.RedTextPosition.HEIGHT_PADDING;
 
+import java.awt.geom.AffineTransform;
+import java.awt.geom.Rectangle2D;
 import java.util.ArrayList;
 import java.util.List;
 import java.util.Locale;
 import java.util.Objects;
+import java.util.regex.Pattern;
 import java.util.stream.Collectors;
 
 import org.apache.pdfbox.text.TextPosition;
@@ -30,6 +33,7 @@ public class TextPositionSequence extends TextBoundingBox implements CharSequenc
     public static final String BOLD_ITALIC = "bold, italic";
     public static final String BOLD = "bold";
     public static final String ITALIC = "italic";
+    public static final Pattern FONT_CLEANER = Pattern.compile(",bold|,italic");
 
     private int page;
 
@@ -154,7 +158,8 @@ public class TextPositionSequence extends TextBoundingBox implements CharSequenc
         if (textPositions.get(0).getFontName() == null) {
             return "none";
         }
-        return textPositions.get(0).getFontName().toLowerCase(Locale.ROOT).replaceAll(",bold", "").replaceAll(",italic", "");
+
+        return FONT_CLEANER.matcher(textPositions.get(0).getFontName().toLowerCase(Locale.ROOT)).replaceAll("");
     }
 
 
@@ -238,5 +243,15 @@ public class TextPositionSequence extends TextBoundingBox implements CharSequenc
         return result;
     }
 
+
+    public void transform(AffineTransform rotateInstance) {
+
+        for (RedTextPosition textPosition : getTextPositions()) {
+            Rectangle2D exactDirAdjCoordinates = rotateInstance.createTransformedShape(textPosition.getBBoxDirAdj()).getBounds2D();
+            textPosition.setBBoxDirAdj(exactDirAdjCoordinates);
+        }
+        calculateBBoxAndHashcode();
+    }
+
 }
 

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

@@ -0,0 +1,34 @@
+package com.knecon.fforesight.service.layoutparser.processor.model.text;
+
+import java.util.Comparator;
+import java.util.HashMap;
+
+public class TextPositionSequenceComparator implements Comparator<TextPositionSequence> {
+
+    private HashMap<TextPositionSequence, TextBlockOnPage> lookup;
+
+
+    public TextPositionSequenceComparator(HashMap<TextPositionSequence, TextBlockOnPage> lookup) {
+
+        this.lookup = lookup;
+    }
+
+
+    @Override
+    public int compare(TextPositionSequence number1, TextPositionSequence number2) {
+
+        int page1 = lookup.get(number1).page().getPageNumber();
+        int page2 = lookup.get(number2).page().getPageNumber();
+
+        if (page1 != page2) {
+            return Integer.compare(page1, page2);
+        }
+
+        if (number1.getY() != number2.getY()) {
+            return Double.compare(number1.getY(), number2.getY());
+        }
+
+        return Integer.compare(Integer.parseInt(number1.toString()), Integer.parseInt(number2.toString()));
+    }
+
+}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/blockification/BlockificationPostprocessingService.java

@@ -7,17 +7,17 @@ import java.util.List;
 import java.util.ListIterator;
 import java.util.Locale;
 
-import org.apache.commons.lang3.StringUtils;
 import org.apache.commons.text.similarity.LevenshteinDistance;
 import org.springframework.stereotype.Service;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.LayoutEngine;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.LayoutEngineProto.LayoutEngine;
 import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationPage;
 import com.knecon.fforesight.service.layoutparser.processor.model.PageBlockType;
 import com.knecon.fforesight.service.layoutparser.processor.model.SectionIdentifier;
 import com.knecon.fforesight.service.layoutparser.processor.model.outline.OutlineObject;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPositionSequence;
+import com.knecon.fforesight.service.layoutparser.processor.utils.TextNormalizationUtilities;
 
 import lombok.Data;
 
@@ -467,7 +467,9 @@ public class BlockificationPostprocessingService {
 
     private static String sanitizeString(String text) {
 
-        return StringUtils.deleteWhitespace(text).toLowerCase(Locale.ROOT);
+        return TextNormalizationUtilities.removeAllWhitespaces(text)//
+                .trim() // sometimes there are trailing empty bytes at the end of the string trim() seems to remove them
+                .toLowerCase(Locale.ENGLISH);
     }
 
 

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

@@ -10,7 +10,6 @@ import com.knecon.fforesight.service.layoutparser.processor.model.Classification
 import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationPage;
 import com.knecon.fforesight.service.layoutparser.processor.model.PageBlockType;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
-import com.knecon.fforesight.service.layoutparser.processor.utils.MarkedContentUtils;
 import com.knecon.fforesight.service.layoutparser.processor.utils.PositionUtils;
 
 import lombok.RequiredArgsConstructor;
@@ -60,26 +59,22 @@ public class ClarifyndClassificationService {
             headlineClassificationService.setLastHeadlineFromOutline(textBlock);
             return;
         }
+        if (textBlock.getClassification() != null && (textBlock.getClassification().equals(PageBlockType.HEADER)
+                                                      || textBlock.getClassification().equals(PageBlockType.FOOTER)
+                                                      || textBlock.getClassification().equals(PageBlockType.TABLE_OF_CONTENTS_ITEM))) {
+            return;
+        }
         if (document.getFontSizeCounter().getMostPopular() == null) {
             textBlock.setClassification(PageBlockType.PARAGRAPH);
             return;
         }
-        if (MarkedContentUtils.intersects(textBlock, page.getMarkedContentBboxPerType(), MarkedContentUtils.HEADER)
-            || PositionUtils.isOverBodyTextFrame(bodyTextFrame, textBlock, page.getRotation()) && (document.getFontSizeCounter().getMostPopular() == null
-                                                                                                   || textBlock.getHighestFontSize() <= document.getFontSizeCounter()
-                .getMostPopular())) {
-            textBlock.setClassification(PageBlockType.PARAGRAPH);
-
-        } else if (MarkedContentUtils.intersects(textBlock, page.getMarkedContentBboxPerType(), MarkedContentUtils.FOOTER)
-                   || PositionUtils.isUnderBodyTextFrame(bodyTextFrame, textBlock, page.getRotation()) && (document.getFontSizeCounter().getMostPopular() == null
-                                                                                                           || textBlock.getHighestFontSize() <= document.getFontSizeCounter()
-                .getMostPopular())) {
-            textBlock.setClassification(PageBlockType.PARAGRAPH);
-        } else if (page.getPageNumber() == 1 //
-                   && (PositionUtils.getHeightDifferenceBetweenChunkWordAndDocumentWord(textBlock, document.getTextHeightCounter().getMostPopular()) > 2.5
-                       && textBlock.getHighestFontSize() > document.getFontSizeCounter().getMostPopular() || page.getTextBlocks().size() == 1)) {
+        if (page.getPageNumber() == 1 //
+            && (PositionUtils.getHeightDifferenceBetweenChunkWordAndDocumentWord(textBlock, document.getTextHeightCounter().getMostPopular()) > 2.5
+                && textBlock.getHighestFontSize() > document.getFontSizeCounter().getMostPopular() || page.getTextBlocks().size() == 1)) {
             if (!Pattern.matches("[0-9]+", textBlock.toString())) {
-                textBlock.setClassification(PageBlockType.TITLE);
+                PageBlockType headlineType = HeadlineClassificationService.headlineClassByFontSize(textBlock, headlineFontSizes);
+                headlineClassificationService.classifyHeadline(textBlock, headlineType);
+                document.setHeadlines(true);
             }
         } else if (textBlock.getMostPopularWordFontSize() > document.getFontSizeCounter().getMostPopular()
                    && PositionUtils.getApproxLineCount(textBlock) < 4.9

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

@@ -0,0 +1,29 @@
+package com.knecon.fforesight.service.layoutparser.processor.services.classification;
+
+import java.util.regex.Pattern;
+
+public class ClassificationPatterns {
+
+    public static final Pattern HEADLINE_WITH_2_IDENTIFER_PATTERN = Pattern.compile("^([1-9]\\d?\\.){1,3}\\d{1,2}\\.?\\s(?:14C)?\\s*[a-z][0-9a-z \\[\\]]{2,50}", Pattern.CASE_INSENSITIVE);
+
+    public static final Pattern HEADLINE_WITH_SINGLE_IDENTIFER_PATTERN = Pattern.compile("^([0-9]\\.)\\s[a-z][0-9a-z \\[\\]]{2,50}", Pattern.CASE_INSENSITIVE);
+
+    public static final Pattern AT_LEAST_3_CHARS_PATTERN = Pattern.compile("\\p{L}{3,}", Pattern.CASE_INSENSITIVE);
+
+    public static final Pattern HEADLINE_PATTERN_WITH_SLASHES = Pattern.compile("^(\\d{1,1}\\.){1,3}\\d{1,2}\\.?\\s[a-z]{1,2}\\/[a-z]{1,2}.*");
+
+    public static final Pattern AMOUNT_PATTERN = Pattern.compile(
+            "^\\s*\\d+(?:\\.\\d+)?\\s*(?:ml|ul|μl|l|ug|μg|g|kg|mg|cm|cm2|cm3|mm|mm2|mm3|km|km2|m|m2|m3|lb|oz|ppm|dpm|days|weeks|months|%|f|ppb)\\b",
+            Pattern.CASE_INSENSITIVE);
+
+
+
+    public static final Pattern TABLE_OR_FIGURE_PATTERN = Pattern.compile(
+            "^\\s*(?:table|continued\\s+table|appendix|figure)\\s+(?:[xvi]+|[a-z0-9]{1,3}(?:\\.[0-9]{1,3})*(?:-[0-9]{1,3})?)\\b",
+            Pattern.CASE_INSENSITIVE);
+
+    public static final Pattern ALPHANUMERIC = Pattern.compile("[a-zA-Z0-9]");
+
+    public static final Pattern NUMERIC = Pattern.compile("[0-9]+");
+
+}

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

@@ -0,0 +1,62 @@
+package com.knecon.fforesight.service.layoutparser.processor.services.classification;
+
+import java.util.Map;
+
+import org.springframework.stereotype.Service;
+
+import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingType;
+import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationDocument;
+import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationPage;
+import com.knecon.fforesight.service.layoutparser.processor.model.PageBlockType;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.services.BodyTextFrameService;
+import com.knecon.fforesight.service.layoutparser.processor.services.blockification.DocuMineBlockificationService;
+
+import lombok.AccessLevel;
+import lombok.RequiredArgsConstructor;
+import lombok.experimental.FieldDefaults;
+import lombok.extern.slf4j.Slf4j;
+
+@Slf4j
+@Service
+@RequiredArgsConstructor
+@FieldDefaults(makeFinal = true, level = AccessLevel.PRIVATE)
+public class ClassificationService {
+
+    DocuMineBlockificationService docuMineBlockificationService;
+    BodyTextFrameService bodyTextFrameService;
+    TableOfContentsClassificationService tableOfContentsClassificationService;
+    RedactManagerClassificationService redactManagerClassificationService;
+    ClarifyndClassificationService clarifyndClassificationService;
+    DocuMineClassificationService docuMineClassificationService;
+    HeaderFooterClassificationService headerFooterClassificationService;
+
+
+    public void classify(ClassificationDocument document, LayoutParsingType layoutParsingType, Map<String, String> identifier) {
+
+        log.info("Calculating BodyTextFrame for {}", identifier);
+        bodyTextFrameService.setBodyTextFrames(document, layoutParsingType);
+        for (ClassificationPage page : document.getPages()) {
+            document.getLayoutDebugLayer().addCleanRulingVisualization(page.getCleanRulings(), page.getPageNumber());
+        }
+        log.info("Classify TextBlocks for {}", identifier);
+
+        headerFooterClassificationService.classifyHeadersAndFooters(document);
+
+        tableOfContentsClassificationService.classifyTableOfContents(document);
+
+        switch (layoutParsingType) {
+            case REDACT_MANAGER, REDACT_MANAGER_PARAGRAPH_DEBUG, REDACT_MANAGER_OLD, CLARIFYND_PARAGRAPH_DEBUG, REDACT_MANAGER_WITHOUT_DUPLICATE_PARAGRAPH ->
+                    redactManagerClassificationService.classifyDocument(document);
+            case DOCUMINE_OLD, DOCUMINE -> docuMineClassificationService.classifyDocument(document);
+            case CLARIFYND -> clarifyndClassificationService.classifyDocument(document);
+        }
+
+        if (layoutParsingType.equals(LayoutParsingType.DOCUMINE_OLD)) {
+            for (ClassificationPage page : document.getPages()) {
+                docuMineBlockificationService.mergeblocks(page, page.getCleanRulings().withoutTextRulings(), 0, 10);
+            }
+        }
+    }
+
+}

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

@@ -1,8 +1,19 @@
 package com.knecon.fforesight.service.layoutparser.processor.services.classification;
 
+import static com.knecon.fforesight.service.layoutparser.processor.services.classification.ClassificationPatterns.ALPHANUMERIC;
+import static com.knecon.fforesight.service.layoutparser.processor.services.classification.ClassificationPatterns.AMOUNT_PATTERN;
+import static com.knecon.fforesight.service.layoutparser.processor.services.classification.ClassificationPatterns.AT_LEAST_3_CHARS_PATTERN;
+import static com.knecon.fforesight.service.layoutparser.processor.services.classification.ClassificationPatterns.HEADLINE_PATTERN_WITH_SLASHES;
+import static com.knecon.fforesight.service.layoutparser.processor.services.classification.ClassificationPatterns.HEADLINE_WITH_2_IDENTIFER_PATTERN;
+import static com.knecon.fforesight.service.layoutparser.processor.services.classification.ClassificationPatterns.HEADLINE_WITH_SINGLE_IDENTIFER_PATTERN;
+import static com.knecon.fforesight.service.layoutparser.processor.services.classification.ClassificationPatterns.TABLE_OR_FIGURE_PATTERN;
+
 import java.util.ArrayList;
+import java.util.Comparator;
+import java.util.Iterator;
 import java.util.List;
 import java.util.Locale;
+import java.util.Map;
 import java.util.regex.Matcher;
 import java.util.regex.Pattern;
 
@@ -12,85 +23,60 @@ import com.knecon.fforesight.service.layoutparser.processor.model.AbstractPageBl
 import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationDocument;
 import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationPage;
 import com.knecon.fforesight.service.layoutparser.processor.model.PageBlockType;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.AbstractBlockOnPage;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.ListIdentifier;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
-import com.knecon.fforesight.service.layoutparser.processor.utils.HeaderFooterDetection;
-import com.knecon.fforesight.service.layoutparser.processor.utils.MarkedContentUtils;
 import com.knecon.fforesight.service.layoutparser.processor.utils.PositionUtils;
 
+import lombok.AccessLevel;
 import lombok.RequiredArgsConstructor;
+import lombok.experimental.FieldDefaults;
 import lombok.extern.slf4j.Slf4j;
 
 @Slf4j
 @Service
 @RequiredArgsConstructor
+@FieldDefaults(makeFinal = true, level = AccessLevel.PRIVATE)
 public class DocuMineClassificationService {
 
-    private static final Pattern HEADLINE_WITH_2_IDENTIFER_PATTERN = Pattern.compile("^([1-9]\\d?\\.){1,3}\\d{1,2}\\.?\\s[a-z][0-9a-z \\[\\]]{2,50}", Pattern.CASE_INSENSITIVE);
-    private static final Pattern HEADLINE_WITH_SINGLE_IDENTIFER_PATTERN = Pattern.compile("^([0-9]\\.)\\s[a-z][0-9a-z \\[\\]]{2,50}", Pattern.CASE_INSENSITIVE);
-    private static final Pattern AT_LEAST_3_CHARS_PATTERN = Pattern.compile("\\p{L}{3,}", Pattern.CASE_INSENSITIVE);
-    private static final Pattern HEADLINE_PATTERN_WITH_SLASHES = Pattern.compile("^(\\d{1,1}\\.){1,3}\\d{1,2}\\.?\\s[a-z]{1,2}\\/[a-z]{1,2}.*");
-    private static final Pattern AMOUNT_PATTERN = Pattern.compile("^\\s*\\d+(?:\\.\\d+)?\\s*(?:ml|l|g|kg|mg|cm|mm|km|m|lb|oz|ppm|%|f)\\b", Pattern.CASE_INSENSITIVE);
-    private static final Pattern TABLE_OR_FIGURE_PATTER = Pattern.compile(
-            "^\\s*(?:table|continued\\s+table|appendix|figure)\\s+(?:[xvi]+|[a-z0-9]{1,3}(?:\\.[0-9]{1,3})*(?:-[0-9]{1,3})?)\\b",
-            Pattern.CASE_INSENSITIVE);
-
     public static final int SEPARATION_THRESHOLD = 10; // if the min distance between a textblock and all its surrounding blocks, the regexes can be more lenient.
     public static final int SURROUNDING_BLOCKS_RADIUS = 3; // number of surrounding blocks before and after the current textblock to be tested
 
+    ListItemClassificationService listItemClassificationService;
+
 
     public void classifyDocument(ClassificationDocument document) {
 
-        List<Double> headlineFontSizes = document.getFontSizeCounter().getValuesInReverseOrder();
-
-        log.debug("Document FontSize counters are: {}", document.getFontSizeCounter().getCountPerValue());
+        List<Double> headlineFontSizes = buildHeadlineFontSizes(document);
+        List<AbstractBlockOnPage> blocks = buildBlocksPerPage(document);
+        log.debug("Headline FontSizes are: {}", headlineFontSizes);
 
         HeadlineClassificationService headlineClassificationService = new HeadlineClassificationService();
 
-        for (ClassificationPage page : document.getPages()) {
-            document.getLayoutDebugLayer().addTextBlockVisualizations(page.getTextBlocks(), page.getPageNumber());
-            classifyPage(headlineClassificationService, page, document, headlineFontSizes);
+        for (int i = 0; i < blocks.size(); i++) {
+            AbstractBlockOnPage block = blocks.get(i);
+            document.getLayoutDebugLayer().addTextBlockVisualizations(block.page().getTextBlocks(), block.page().getPageNumber());
+            classifyBlock(headlineClassificationService, i, blocks, document, headlineFontSizes);
         }
-    }
 
-
-    private void classifyPage(HeadlineClassificationService headlineClassificationService,
-                              ClassificationPage page,
-                              ClassificationDocument document,
-                              List<Double> headlineFontSizes) {
-
-        List<AbstractPageBlock> textBlocks = page.getTextBlocks();
-        for (int i = 0; i < textBlocks.size(); i++) {
-            AbstractPageBlock textBlock = textBlocks.get(i);
-            if (textBlock instanceof TextPageBlock) {
-                List<AbstractPageBlock> surroundingBlocks = getSurroundingBlocks(i, textBlocks);
-                classifyBlock(headlineClassificationService, (TextPageBlock) textBlock, surroundingBlocks, page, document, headlineFontSizes);
-            }
-        }
-    }
-
-
-    private List<AbstractPageBlock> getSurroundingBlocks(int originalIndex, List<AbstractPageBlock> textBlocks) {
-
-        int start = Math.max(originalIndex - SURROUNDING_BLOCKS_RADIUS, 0);
-        int end = Math.min(originalIndex + SURROUNDING_BLOCKS_RADIUS, textBlocks.size());
-        List<AbstractPageBlock> surroundingBlocks = new ArrayList<>(2 * SURROUNDING_BLOCKS_RADIUS);
-        for (int i = start; i < end; i++) {
-            if (i == originalIndex) {
-                continue;
-            }
-            surroundingBlocks.add(textBlocks.get(i));
-        }
-        return surroundingBlocks;
     }
 
 
     private void classifyBlock(HeadlineClassificationService headlineClassificationService,
-                               TextPageBlock textBlock,
-                               List<AbstractPageBlock> surroundingBlocks,
-                               ClassificationPage page,
+                               int currentIndex,
+                               List<AbstractBlockOnPage> allBlocks,
                                ClassificationDocument document,
                                List<Double> headlineFontSizes) {
 
+        TextPageBlock textBlock;
+        if (allBlocks.get(currentIndex).block() instanceof TextPageBlock block) {
+            textBlock = block;
+        } else {
+            return;
+        }
+        ClassificationPage page = allBlocks.get(currentIndex).page();
+        List<AbstractPageBlock> surroundingBlocks = getSurroundingBlocksOnPage(currentIndex, allBlocks);
+
         log.debug("headlineFontSizes: {}", headlineFontSizes);
         var bodyTextFrame = page.getBodyTextFrame();
 
@@ -98,72 +84,57 @@ public class DocuMineClassificationService {
         Matcher atLeast3Matcher = AT_LEAST_3_CHARS_PATTERN.matcher(textBlock.toString());
         Matcher headlineWithSlashesMatcher = HEADLINE_PATTERN_WITH_SLASHES.matcher(textBlock.toString());
         Matcher amountMatcher = AMOUNT_PATTERN.matcher(textBlock.toString());
-        Matcher tableOrFigureMatcher = TABLE_OR_FIGURE_PATTER.matcher(textBlock.toString());
+        Matcher tableOrFigureMatcher = TABLE_OR_FIGURE_PATTERN.matcher(textBlock.toString());
         Matcher headlineWithSingleIdentifierMatcher = HEADLINE_WITH_SINGLE_IDENTIFER_PATTERN.matcher(textBlock.toString());
         boolean isAtLeast3Characters = atLeast3Matcher.reset().find();
-        boolean isTocItem = textBlock.getText().contains("..............");
         boolean headlineWithSlashesMatches = headlineWithSlashesMatcher.reset().matches();
         boolean isAmount = amountMatcher.reset().find();
         int charCount = countChars(textBlock);
 
         boolean enoughChars = charCount > textBlock.getText().length() * 0.5;
 
+        List<ListIdentifier> listIdentifiers = listItemClassificationService.findConfirmedListIdentifiers(currentIndex, allBlocks);
+        document.getLayoutDebugLayer().addListIdentifiers(listIdentifiers);
+
         if (textBlock.getClassification() != null && textBlock.getClassification().isHeadline()) {
             headlineClassificationService.setLastHeadlineFromOutline(textBlock);
             return;
         }
+        if (textBlock.getClassification() != null && (textBlock.getClassification().equals(PageBlockType.HEADER)
+                                                      || textBlock.getClassification().equals(PageBlockType.FOOTER)
+                                                      || textBlock.getClassification().equals(PageBlockType.TABLE_OF_CONTENTS_ITEM))) {
+            return;
+        }
         if (document.getFontSizeCounter().getMostPopular() == null) {
             textBlock.setClassification(PageBlockType.PARAGRAPH);
             return;
         }
-        if (MarkedContentUtils.intersects(textBlock, page.getMarkedContentBboxPerType(), MarkedContentUtils.HEADER) //
-            || (PositionUtils.isOverBodyTextFrame(bodyTextFrame, textBlock, page.getRotation()) //
-                && (document.getFontSizeCounter().getMostPopular() == null //
-                    || textBlock.getHighestFontSize() <= document.getFontSizeCounter().getMostPopular()))) {
-            textBlock.setClassification(PageBlockType.HEADER);
-
-        } else if (MarkedContentUtils.intersects(textBlock, page.getMarkedContentBboxPerType(), MarkedContentUtils.FOOTER)
-                   || (PositionUtils.isUnderBodyTextFrame(bodyTextFrame,
-                                                          textBlock,
-                                                          page.getRotation())
-                       && (document.getFontSizeCounter().getMostPopular()
-                           == null
-                           || textBlock.getHighestFontSize()
-                              <= document.getFontSizeCounter()
-                                      .getMostPopular()))
-                   || HeaderFooterDetection.isLikelyFooter(textBlock, document, page)) {
-            textBlock.setClassification(PageBlockType.FOOTER);
-        } else if (page.getPageNumber() == 1 && (PositionUtils.getHeightDifferenceBetweenChunkWordAndDocumentWord(textBlock, document.getTextHeightCounter().getMostPopular()) > 2.5
-                                                 && textBlock.getHighestFontSize() > document.getFontSizeCounter().getMostPopular() || page.getTextBlocks().size() == 1)) {
-            if (!Pattern.matches("[0-9]+", textBlock.toString())) {
-                textBlock.setClassification(PageBlockType.TITLE);
-            }
-        } else if (textBlock.getText().length() > 5
-                   && (textBlock.getMostPopularWordHeight() > document.getTextHeightCounter().getMostPopular()
-                       || textBlock.getMostPopularWordFontSize() > document.getFontSizeCounter().getMostPopular())
-                   && PositionUtils.getApproxLineCount(textBlock) < 5.9
-                   && ((textBlock.getMostPopularWordStyle().contains("bold") || textBlock.isUnderlined())//
-                       && Character.isDigit(textBlock.toString().charAt(0)) //
-                       && isAtLeast3Characters //
-                       && !textBlock.toString().contains(":") //
-                       || textBlock.toString().startsWith("APPENDIX") //
-                       || textBlock.toString().startsWith("FIGURE") //
-                       || textBlock.toString().startsWith("Continued TABLE") //
-                       || textBlock.toString().startsWith("TABLE"))
-                   && !textBlock.toString().endsWith(":")
-                   && isAtLeast3Characters
-                   && !isTocItem
-                   && !isAmount
-                   && enoughChars) {
+        if (textBlock.getText().length() > 5
+            && greaterOrEqualFontThanDocumentAverage(textBlock, document)
+            && PositionUtils.getApproxLineCount(textBlock) < 5.9
+            && ((textBlock.getMostPopularWordStyle().contains("bold") || textBlock.isUnderlined())//
+                && Character.isDigit(textBlock.toString().charAt(0)) //
+                && isAtLeast3Characters //
+                && !textBlock.toString().contains(":") //
+                || textBlock.toString().startsWith("APPENDIX") //
+                || textBlock.toString().startsWith("FIGURE") //
+                || textBlock.toString().startsWith("Continued TABLE") //
+                || textBlock.toString().startsWith("TABLE"))
+            && !textBlock.toString().endsWith(":")
+            && isAtLeast3Characters
+            && !isAmount
+            && enoughChars) {
 
             setAsHeadline(headlineClassificationService, textBlock, document, headlineFontSizes);
         } else if (isAllCaps(textBlock)
+                   && ALPHANUMERIC.matcher(Character.toString(textBlock.getText().charAt(0))).matches()
+                   && hasSeparation(textBlock, surroundingBlocks)
                    && textBlock.getText().length() > 5
                    && isAtLeast3Characters
                    && !isAmount
                    && enoughChars
                    && !textBlock.toString().contains(":")
-                   && !textBlock.toString().startsWith("(")
+                   && !textBlock.toString().endsWith(".")
                    && PositionUtils.getApproxLineCount(textBlock) < 2.9) {
 
             setAsHeadline(headlineClassificationService, textBlock, document, headlineFontSizes);
@@ -171,19 +142,20 @@ public class DocuMineClassificationService {
                    && PositionUtils.getApproxLineCount(textBlock) < 2.9
                    && isAtLeast3Characters
                    && !headlineWithSlashesMatches
-                   && !isAmount
-                   && !isTocItem) {
+                   && !isAmount) {
 
             setAsHeadline(headlineClassificationService, textBlock, document, headlineFontSizes);
-        } else if (!isTocItem
-                   && hasSeparation(textBlock, surroundingBlocks)
-                   && greaterOrEqualThanFontPageAverage(textBlock, page)
-                   && PositionUtils.getApproxLineCount(textBlock) < 2.9
-                   && (tableOrFigureMatcher.reset().find() || headlineWithSingleIdentifierMatcher.reset().find())
-                   && !isAmount
+        } else if (hasSeparation(textBlock, surroundingBlocks)//
+                   && greaterOrEqualFontThanPageAverage(textBlock, page)//
+                   && PositionUtils.getApproxLineCount(textBlock) < 2.9//
+                   && (tableOrFigureMatcher.reset().find() || (headlineWithSingleIdentifierMatcher.reset().find() && listIdentifiers.isEmpty())) //
+                   && !isAmount//
                    && !headlineWithSlashesMatches) {
 
             setAsHeadline(headlineClassificationService, textBlock, document, headlineFontSizes);
+        } else if (!listIdentifiers.isEmpty()) {
+
+            textBlock.setClassification(PageBlockType.LIST_ITEM);
         } else if (PositionUtils.isWithinBodyTextFrame(bodyTextFrame, textBlock)
                    && textBlock.getMostPopularWordFontSize() == document.getFontSizeCounter().getMostPopular()
                    && textBlock.getMostPopularWordStyle().equals("bold")
@@ -222,13 +194,20 @@ public class DocuMineClassificationService {
     }
 
 
-    private static boolean greaterOrEqualThanFontPageAverage(TextPageBlock textBlock, ClassificationPage page) {
+    private static boolean greaterOrEqualFontThanPageAverage(TextPageBlock textBlock, ClassificationPage page) {
 
         return textBlock.getMostPopularWordHeight() >= page.getTextHeightCounter().getMostPopular() //
                || textBlock.getMostPopularWordFontSize() >= page.getFontSizeCounter().getMostPopular();
     }
 
 
+    private static boolean greaterOrEqualFontThanDocumentAverage(TextPageBlock textBlock, ClassificationDocument document) {
+
+        return textBlock.getMostPopularWordHeight() > document.getTextHeightCounter().getMostPopular() //
+               || textBlock.getMostPopularWordFontSize() > document.getFontSizeCounter().getMostPopular();
+    }
+
+
     private static boolean isAllCaps(TextPageBlock textBlock) {
 
         return textBlock.toString().equals(textBlock.toString().toUpperCase(Locale.ROOT));
@@ -246,8 +225,7 @@ public class DocuMineClassificationService {
 
         return surroundingBlocks.stream()
                 .mapToDouble(surroundingBlock -> calculateSeparation(textBlock, surroundingBlock))
-                .min()
-                .orElse(Double.MAX_VALUE);
+                .min().orElse(Double.MAX_VALUE);
     }
 
 
@@ -267,6 +245,92 @@ public class DocuMineClassificationService {
         document.setHeadlines(true);
     }
 
+
+    private List<AbstractBlockOnPage> buildBlocksPerPage(ClassificationDocument document) {
+
+        List<AbstractBlockOnPage> blocks = new ArrayList<>();
+        for (ClassificationPage page : document.getPages()) {
+            for (AbstractPageBlock abstractPageBlock : page.getTextBlocks()) {
+                if (abstractPageBlock instanceof TextPageBlock textBlock) {
+                    if (textBlock.getClassification() != null && (textBlock.getClassification().equals(PageBlockType.HEADER) //
+                                                                  || textBlock.getClassification().equals(PageBlockType.FOOTER))) {
+                        continue;
+                    }
+                    blocks.add(new AbstractBlockOnPage(textBlock, page));
+                }
+            }
+        }
+        return blocks;
+    }
+
+
+    private static List<Double> buildHeadlineFontSizes(ClassificationDocument document) {
+
+        if (document.getFontSizeCounter().getCountPerValue().size() <= 6) {
+            return document.getFontSizeCounter().getValuesInReverseOrder();
+        }
+
+        List<Map.Entry<Double, Integer>> sortedEntries = new ArrayList<>(document.getFontSizeCounter().getCountPerValue().entrySet());
+        sortedEntries.sort(Map.Entry.comparingByKey());
+
+        int totalCount = sortedEntries.stream()
+                .mapToInt(Map.Entry::getValue).sum();
+
+        int cumulativeCount = 0;
+        Iterator<Map.Entry<Double, Integer>> iterator = sortedEntries.iterator();
+        while (iterator.hasNext()) {
+            Map.Entry<Double, Integer> entry = iterator.next();
+            cumulativeCount += entry.getValue();
+            if (cumulativeCount > totalCount * 0.3) {
+                break; // We've filtered the bottom 30%, so stop.
+            }
+            iterator.remove();
+        }
+
+        if (sortedEntries.size() < 6) {
+            return document.getFontSizeCounter().getValuesInReverseOrder();
+        }
+        int clusterSize = Math.max(1, sortedEntries.size() / 6);
+
+        List<List<Double>> clusters = new ArrayList<>();
+        for (int i = 0; i < 6; i++) {
+            clusters.add(new ArrayList<>());
+        }
+
+        for (int i = 0; i < sortedEntries.size(); i++) {
+            int clusterIndex = Math.min(i / clusterSize, 5);
+            clusters.get(clusterIndex).add(sortedEntries.get(i).getKey());
+        }
+
+        return clusters.stream()
+                .map(cluster -> cluster.stream()
+                        .mapToDouble(d -> d).average()
+                        .orElseThrow())
+                .sorted(Comparator.reverseOrder())
+                .toList();
+    }
+
+
+    private List<AbstractPageBlock> getSurroundingBlocksOnPage(int originalIndex, List<AbstractBlockOnPage> textBlocks) {
+
+        int start = Math.max(originalIndex - SURROUNDING_BLOCKS_RADIUS, 0);
+        int end = Math.min(originalIndex + SURROUNDING_BLOCKS_RADIUS, textBlocks.size());
+        List<AbstractPageBlock> surroundingBlocks = new ArrayList<>(2 * SURROUNDING_BLOCKS_RADIUS);
+        for (int i = start; i < end; i++) {
+            if (i == originalIndex) {
+                continue;
+            }
+            if (textBlocks.get(i).block().getText().length() <= 1) {
+                continue;
+            }
+            if (!textBlocks.get(i).page().equals(textBlocks.get(originalIndex).page())) {
+                continue;
+            }
+            surroundingBlocks.add(textBlocks.get(i).block());
+        }
+        return surroundingBlocks;
+    }
+
 }
 
 

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

@@ -0,0 +1,55 @@
+package com.knecon.fforesight.service.layoutparser.processor.services.classification;
+
+import org.springframework.stereotype.Service;
+
+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;
+import com.knecon.fforesight.service.layoutparser.processor.model.PageBlockType;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.utils.MarkedContentUtils;
+import com.knecon.fforesight.service.layoutparser.processor.utils.PositionUtils;
+
+import lombok.AccessLevel;
+import lombok.RequiredArgsConstructor;
+import lombok.experimental.FieldDefaults;
+
+@Service
+@RequiredArgsConstructor
+@FieldDefaults(makeFinal = true, level = AccessLevel.PRIVATE)
+public class HeaderFooterClassificationService {
+
+    public void classifyHeadersAndFooters(ClassificationDocument document) {
+
+        for (ClassificationPage page : document.getPages()) {
+            for (AbstractPageBlock pageBlock : page.getTextBlocks()) {
+                if (pageBlock instanceof TextPageBlock textBlock) {
+                    classifyBlock(document, page, textBlock);
+                }
+            }
+        }
+
+    }
+
+
+    private static void classifyBlock(ClassificationDocument document, ClassificationPage page, TextPageBlock textBlock) {
+
+        if (MarkedContentUtils.intersects(textBlock, page.getMarkedContentBboxPerType(), MarkedContentUtils.HEADER)
+            || PositionUtils.isOverBodyTextFrame(page.getBodyTextFrame(), textBlock, page.getRotation()) && smallerFontThanDocAverage(document, textBlock)) {
+
+            textBlock.setClassification(PageBlockType.HEADER);
+
+        } else if (MarkedContentUtils.intersects(textBlock, page.getMarkedContentBboxPerType(), MarkedContentUtils.FOOTER)
+                   || PositionUtils.isUnderBodyTextFrame(page.getBodyTextFrame(), textBlock, page.getRotation()) && smallerFontThanDocAverage(document, textBlock)) {
+
+            textBlock.setClassification(PageBlockType.FOOTER);
+        }
+    }
+
+
+    private static boolean smallerFontThanDocAverage(ClassificationDocument document, TextPageBlock textBlock) {
+
+        return document.getFontSizeCounter().getMostPopular() == null || textBlock.getHighestFontSize() <= document.getFontSizeCounter().getMostPopular();
+    }
+
+}

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

@@ -73,13 +73,18 @@ public class HeadlineClassificationService {
 
     public static PageBlockType headlineClassByFontSize(TextPageBlock textBlock, List<Double> fontSizeGroups) {
 
-        PageBlockType headlineType = PageBlockType.H1;
-        for (int i = 1; i <= fontSizeGroups.size(); i++) {
-            if (textBlock.getMostPopularWordFontSize() == fontSizeGroups.get(i - 1)) {
-                headlineType = PageBlockType.getHeadlineType(i);
+        List<Double> distances = fontSizeGroups.stream()
+                .map(fontSize -> Math.abs(fontSize - textBlock.getMostPopularWordFontSize()))
+                .toList();
+        double min = Double.MAX_VALUE;
+        int argMin = -1;
+        for (int i = 0; i < distances.size(); i++) {
+            if (distances.get(i) < min) {
+                min = distances.get(i);
+                argMin = i;
             }
         }
-        return headlineType;
+        return PageBlockType.getHeadlineType(argMin);
     }
 
 }

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

@@ -0,0 +1,99 @@
+package com.knecon.fforesight.service.layoutparser.processor.services.classification;
+
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.Collections;
+import java.util.LinkedList;
+import java.util.List;
+
+import org.springframework.stereotype.Service;
+
+import com.knecon.fforesight.service.layoutparser.processor.model.text.AbstractBlockOnPage;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.ListIdentifier;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPositionSequence;
+
+@Service
+public class ListItemClassificationService {
+
+    public static final int LIST_IDENTIFIER_SEARCH_RADIUS = 3;
+
+
+    public List<ListIdentifier> findConfirmedListIdentifiers(int currentIndex, List<AbstractBlockOnPage> allBlocks) {
+
+        List<ListIdentifier> listIdentifiers = extractListIdentifiers(allBlocks.get(currentIndex));
+        if (listIdentifiers.isEmpty()) {
+            return Collections.emptyList();
+        }
+        if (listIdentifiers.size() > 1 && ListIdentifier.isInOrder(listIdentifiers)) {
+            return listIdentifiers;
+        }
+
+        int start = Math.max(0, currentIndex - LIST_IDENTIFIER_SEARCH_RADIUS);
+        int end = Math.min(allBlocks.size(), currentIndex + LIST_IDENTIFIER_SEARCH_RADIUS);
+
+        List<ListIdentifier> identifiersBehind = new ArrayList<>();
+        if (start < currentIndex) {
+            identifiersBehind.addAll(allBlocks.subList(start, currentIndex)
+                                             .stream()
+                                             .map(this::extractListIdentifiers)
+                                             .flatMap(Collection::stream)
+                                             .toList());
+        }
+        if (!identifiersBehind.isEmpty()) {
+            listIdentifiers.add(0, identifiersBehind.get(identifiersBehind.size() - 1));
+            if (ListIdentifier.isInOrder(listIdentifiers)) {
+                return listIdentifiers;
+            }
+            listIdentifiers.remove(0);
+        }
+        List<ListIdentifier> identifiersAhead = new ArrayList<>();
+        if (currentIndex + 1 < end) {
+            identifiersAhead.addAll(allBlocks.subList(currentIndex + 1, end)
+                                            .stream()
+                                            .map(this::extractListIdentifiers)
+                                            .flatMap(Collection::stream)
+                                            .toList());
+        }
+        if (!identifiersAhead.isEmpty()) {
+            listIdentifiers.add(identifiersAhead.get(0));
+            if (ListIdentifier.isInOrder(listIdentifiers)) {
+                return listIdentifiers;
+            }
+            listIdentifiers.remove(listIdentifiers.size() - 1);
+        }
+        return Collections.emptyList();
+
+    }
+
+
+    private List<ListIdentifier> extractListIdentifiers(AbstractBlockOnPage block) {
+
+        List<ListIdentifier> result = new LinkedList<>();
+        if (block.block() instanceof TextPageBlock textBlock) {
+            List<TextPositionSequence> sequences = textBlock.getSequences();
+            for (int i = 0; i < sequences.size(); i++) {
+
+                if (i != 0 && sequences.get(i - 1).getXDirAdj() < sequences.get(i).getXDirAdj()) {
+                    // is not the start of a line, continue
+                    continue;
+                }
+
+                TextPositionSequence sequence = sequences.get(i);
+                List<TextPositionSequence> wordsAtStartOfLine = new ArrayList<>(3);
+                int end = Math.min(sequences.size(), i + 3);
+                for (int j = i; j < end; j++) {
+                    if (sequences.get(j).intersectsYDirAdj(sequence, 2)) {
+                        wordsAtStartOfLine.add(sequences.get(j));
+                    } else {
+                        break;
+                    }
+                }
+
+                ListIdentifier.parse(wordsAtStartOfLine, block.page().getPageNumber()).ifPresent(result::add);
+            }
+        }
+        return result;
+    }
+
+}

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

@@ -61,6 +61,15 @@ public class RedactManagerClassificationService {
             headlineClassificationService.setLastHeadlineFromOutline(textBlock);
             return;
         }
+        if (textBlock.getClassification() != null && (textBlock.getClassification().equals(PageBlockType.HEADER)
+                                                      || textBlock.getClassification().equals(PageBlockType.FOOTER)
+                                                      || textBlock.getClassification().equals(PageBlockType.TABLE_OF_CONTENTS_ITEM))) {
+            return;
+        }
+        if (document.getFontSizeCounter().getMostPopular() == null) {
+            textBlock.setClassification(PageBlockType.PARAGRAPH);
+            return;
+        }
         if (document.getFontSizeCounter().getMostPopular() == null) {
             textBlock.setClassification(PageBlockType.PARAGRAPH);
             return;
@@ -73,24 +82,13 @@ public class RedactManagerClassificationService {
             textBlock.setClassification(PageBlockType.PARAGRAPH);
             return;
         }
-        if (MarkedContentUtils.intersects(textBlock, page.getMarkedContentBboxPerType(), MarkedContentUtils.HEADER)
-            || PositionUtils.isOverBodyTextFrame(bodyTextFrame, textBlock, page.getRotation()) && (document.getFontSizeCounter().getMostPopular() == null
-                                                                                                   || textBlock.getHighestFontSize() <= document.getFontSizeCounter()
-                .getMostPopular())) {
-
-            textBlock.setClassification(PageBlockType.HEADER);
-
-        } else if (MarkedContentUtils.intersects(textBlock, page.getMarkedContentBboxPerType(), MarkedContentUtils.FOOTER)
-                   || PositionUtils.isUnderBodyTextFrame(bodyTextFrame, textBlock, page.getRotation()) && (document.getFontSizeCounter().getMostPopular() == null
-                                                                                                           || textBlock.getHighestFontSize() <= document.getFontSizeCounter()
-                .getMostPopular())) {
-
-            textBlock.setClassification(PageBlockType.FOOTER);
-        } else if (page.getPageNumber() == 1 && (PositionUtils.getHeightDifferenceBetweenChunkWordAndDocumentWord(textBlock, document.getTextHeightCounter().getMostPopular()) > 2.5
-                                                 && textBlock.getHighestFontSize() > document.getFontSizeCounter().getMostPopular() || page.getTextBlocks().size() == 1)) {
+        if (page.getPageNumber() == 1 && (PositionUtils.getHeightDifferenceBetweenChunkWordAndDocumentWord(textBlock, document.getTextHeightCounter().getMostPopular()) > 2.5
+                                          && textBlock.getHighestFontSize() > document.getFontSizeCounter().getMostPopular() || page.getTextBlocks().size() == 1)) {
 
             if (!Pattern.matches("[0-9]+", textBlock.toString())) {
-                textBlock.setClassification(PageBlockType.TITLE);
+                PageBlockType headlineType = HeadlineClassificationService.headlineClassByFontSize(textBlock, headlineFontSizes);
+                headlineClassificationService.classifyHeadline(textBlock, headlineType);
+                document.setHeadlines(true);
             }
         } else if (textBlock.getMostPopularWordFontSize() > document.getFontSizeCounter().getMostPopular()
                    && PositionUtils.getApproxLineCount(textBlock) < 4.9

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

@@ -0,0 +1,370 @@
+package com.knecon.fforesight.service.layoutparser.processor.services.classification;
+
+import static com.knecon.fforesight.service.layoutparser.processor.services.classification.ClassificationPatterns.AMOUNT_PATTERN;
+import static com.knecon.fforesight.service.layoutparser.processor.services.classification.ClassificationPatterns.NUMERIC;
+
+import java.util.ArrayList;
+import java.util.Collection;
+import java.util.Collections;
+import java.util.Comparator;
+import java.util.HashMap;
+import java.util.HashSet;
+import java.util.LinkedList;
+import java.util.List;
+import java.util.Locale;
+import java.util.Map;
+import java.util.Set;
+import java.util.stream.Collectors;
+
+import org.springframework.stereotype.Service;
+
+import com.knecon.fforesight.service.layoutparser.processor.docstrum.model.UnionFind;
+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;
+import com.knecon.fforesight.service.layoutparser.processor.model.PageBlockType;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.TextBlockOnPage;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPositionSequence;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPositionSequenceComparator;
+import com.knecon.fforesight.service.layoutparser.processor.utils.TextNormalizationUtilities;
+import com.knecon.fforesight.service.layoutparser.processor.visualization.LayoutDebugLayer;
+
+import lombok.extern.slf4j.Slf4j;
+
+@Slf4j
+@Service
+public class TableOfContentsClassificationService {
+
+    private static final int MAX_PAGE_COUNT = 10; // maximum length of a toc to avoid runaway classification
+    private static final int SURROUNDING_BLOCKS_RADIUS = 10; // number of blocks to look ahead
+    private static final int MINIMUM_MATCHES = 2; // minimum cluster size
+    public static final int INTERSECTION_TOLERANCE = 2; // equality threshold for x intersection
+    public static final int DENSITY_THRESHOLD_COUNT = 10; // describes the minimum density, at least this many entries per page height are required
+
+
+    @SuppressWarnings("checkstyle:ModifiedControlVariable")
+    public void classifyTableOfContents(ClassificationDocument document) {
+
+        List<TextBlockOnPage> textBlocks = buildBlocksPerPage(document);
+
+        for (int i = 0; i < textBlocks.size(); i++) {
+            TextBlockOnPage textBlock = textBlocks.get(i);
+
+            if (!isTOCHeadline(textBlock)) {
+                continue;
+            }
+
+            int end = identifyTOCItems(i + 1, textBlocks, document);
+
+            if (end > i + 1) {
+                if (textBlock.textBlock().getClassification() == null) {
+                    textBlock.textBlock().setClassification(PageBlockType.H1);
+                }
+                i = end;
+            }
+        }
+    }
+
+
+    private int identifyTOCItems(int start, List<TextBlockOnPage> textBlocks, ClassificationDocument document) {
+
+        ClassificationPage startPage = textBlocks.get(start).page();
+        List<TextBlockOnPage> initialLookAhead = textBlocks.subList(start, Math.min(start + SURROUNDING_BLOCKS_RADIUS, textBlocks.size()));
+        HashMap<TextPositionSequence, TextBlockOnPage> lookup = new HashMap<>();
+        List<TextPositionSequence> numbers = extractNumbers(initialLookAhead, lookup, document.getPages().size());
+        TocNumberFinder tocNumberFinder = new TocNumberFinder(numbers, lookup);
+
+        int lastCandidate = start;
+        for (int i = start; i < Math.min(lastCandidate + SURROUNDING_BLOCKS_RADIUS, textBlocks.size()); i++) {
+
+            TextBlockOnPage textBlockOnPage = textBlocks.get(i);
+            if (textBlockOnPage.page().getPageNumber() - MAX_PAGE_COUNT > startPage.getPageNumber()) {
+                break;
+            }
+
+            if (textBlockOnPage.textBlock().getClassification() != null //
+                && textBlockOnPage.textBlock().getClassification().isHeadline() //
+                && !(textBlockOnPage.textBlock().getText().startsWith("TABLES") //
+                     || textBlockOnPage.textBlock().getText().startsWith("APPENDICES") //
+                     || textBlockOnPage.textBlock().getText().startsWith("FIGURES"))) {
+                log.debug("hit an outline headline, stop immediately.");
+                lastCandidate = i - 1;
+                break;
+            }
+
+            List<TextPositionSequence> numbersFromBlock = extractNumbers(textBlockOnPage, lookup, document.getPages().size());
+
+            List<TextPositionSequence> currentRightmostCluster = tocNumberFinder.getCurrentRightmostCluster();
+
+            if (currentRightmostCluster.size() < MINIMUM_MATCHES) {
+                log.debug("No numbers indicating a table of contents here.");
+                return start;
+            }
+
+            if (anyIntersection(currentRightmostCluster, numbersFromBlock, lookup)) {
+                lastCandidate = i;
+                numbersFromBlock.forEach(tocNumberFinder::add);
+            }
+        }
+
+        addVisualization(document.getLayoutDebugLayer(), tocNumberFinder, lookup);
+
+        Set<TextBlockOnPage> blocksWithNumberInCluster = tocNumberFinder.getCurrentRightmostCluster()
+                .stream()
+                .map(lookup::get)
+                .collect(Collectors.toSet());
+
+        int lastConfirmed = start;
+        for (int i = start; i < lastCandidate + 1; i++) {
+            TextBlockOnPage textBlockOnPage = textBlocks.get(i);
+            if (blocksWithNumberInCluster.contains(textBlockOnPage)) {
+                lastConfirmed = i;
+            }
+        }
+
+        textBlocks.subList(start, lastConfirmed + 1)
+                .stream()
+                .filter(block -> (block.textBlock().getClassification() == null || !block.textBlock().getClassification().isHeadline()))
+                .forEach(textBlockOnPage -> textBlockOnPage.textBlock().setClassification(PageBlockType.TABLE_OF_CONTENTS_ITEM));
+
+        return lastCandidate;
+    }
+
+
+    private static void addVisualization(LayoutDebugLayer layoutDebugLayer, TocNumberFinder tocNumberFinder, Map<TextPositionSequence, TextBlockOnPage> lookup) {
+
+        tocNumberFinder.getCurrentRightmostCluster()
+                .stream()
+                .collect(Collectors.groupingBy(key -> lookup.get(key).page().getPageNumber()))
+                .forEach((pageNumber, number) -> layoutDebugLayer.addTocPages(number, pageNumber));
+    }
+
+
+    private static boolean anyIntersection(Collection<TextPositionSequence> numbers1,
+                                           Collection<TextPositionSequence> numbers2,
+                                           Map<TextPositionSequence, TextBlockOnPage> lookup) {
+
+        return numbers1.stream()
+                .anyMatch(numberFromCluster -> numbers2.stream()
+                        .anyMatch(numberFromBlock -> matches(numberFromBlock, numberFromCluster, lookup)));
+    }
+
+
+    private static List<TextPositionSequence> extractNumbers(List<TextBlockOnPage> textBlocks, Map<TextPositionSequence, TextBlockOnPage> lookup, int numberOfPages) {
+
+        List<TextPositionSequence> blocks = new LinkedList<>();
+        for (TextBlockOnPage textBlock : textBlocks) {
+            blocks.addAll(extractNumbers(textBlock, lookup, numberOfPages));
+        }
+        return blocks;
+    }
+
+
+    private static List<TextPositionSequence> extractNumbers(TextBlockOnPage textBlock, Map<TextPositionSequence, TextBlockOnPage> lookup, int numberOfPages) {
+
+        List<TextPositionSequence> blocks = new LinkedList<>();
+        TextPageBlock block = textBlock.textBlock();
+        List<TextPositionSequence> sequences = block.getSequences();
+        for (int i = 0; i < sequences.size(); i++) {
+
+            TextPositionSequence word = sequences.get(i);
+
+            if (!NUMERIC.matcher(word).matches()) {
+                continue;
+            }
+
+            if (AMOUNT_PATTERN.matcher(getSurroundingString(i, sequences)).matches()) {
+                continue;
+            }
+
+            try {
+                int pageNumber = Integer.parseInt(word.toString());
+                if (0 >= pageNumber || pageNumber > numberOfPages) {
+                    continue;
+                }
+                lookup.put(word, textBlock);
+                blocks.add(word);
+            } catch (NumberFormatException e) {
+                log.debug("That wasn't a number! Should not happen, due to numeric check beforehand.");
+            }
+        }
+        return blocks;
+    }
+
+
+    private static CharSequence getSurroundingString(int i, List<TextPositionSequence> sequences) {
+
+        int end = Math.min(i + 5, sequences.size());
+        return sequences.subList(i, end)
+                .stream()
+                .map(TextPositionSequence::toString)
+                .collect(Collectors.joining(" "));
+    }
+
+
+    private static boolean matches(TextPositionSequence number1, TextPositionSequence number2, Map<TextPositionSequence, TextBlockOnPage> lookup) {
+
+        if (number1.getDir() != number2.getDir()) {
+            return false;
+        }
+
+        return number1.intersectsXDirAdj(number2, INTERSECTION_TOLERANCE);
+    }
+
+
+    private boolean isTOCHeadline(TextBlockOnPage textBlock) {
+
+        if (textBlock.textBlock().getText().length() > 50) {
+            return false;
+        }
+        String text = TextNormalizationUtilities.removeAllWhitespaces(textBlock.textBlock().getText().toLowerCase(Locale.ENGLISH));
+        return (text.contains("content") && text.length() < "content".length() + 6) //
+               || (text.contains("tableofcontent") && text.length() < "tableofcontent".length() + DENSITY_THRESHOLD_COUNT)//
+               || text.equals("tables")//
+               || text.equals("appendices")//
+               || text.equals("figures");
+    }
+
+
+    private List<TextBlockOnPage> buildBlocksPerPage(ClassificationDocument document) {
+
+        List<TextBlockOnPage> blocks = new ArrayList<>();
+        for (ClassificationPage page : document.getPages()) {
+            for (AbstractPageBlock abstractPageBlock : page.getTextBlocks()) {
+                if (abstractPageBlock instanceof TextPageBlock textBlock) {
+                    if (textBlock.getClassification() != null && (textBlock.getClassification().equals(PageBlockType.HEADER) //
+                                                                  || textBlock.getClassification().equals(PageBlockType.FOOTER))) {
+                        continue;
+                    }
+                    blocks.add(new TextBlockOnPage(textBlock, page));
+                }
+            }
+        }
+        return blocks;
+    }
+
+
+    private static class TocNumberFinder {
+
+        final UnionFind<TextPositionSequence> numberClusters;
+        final HashMap<TextPositionSequence, TextBlockOnPage> lookup;
+
+
+        TocNumberFinder(List<TextPositionSequence> blocks, HashMap<TextPositionSequence, TextBlockOnPage> lookup) {
+
+            this.numberClusters = new UnionFind<>(new HashSet<>(blocks));
+            for (int i = 0; i < blocks.size(); i++) {
+                for (int j = i + 1; j < blocks.size(); j++) {
+                    if (matches(blocks.get(i), blocks.get(j), lookup)) {
+                        numberClusters.union(blocks.get(i), blocks.get(j));
+                    }
+                }
+            }
+            this.lookup = lookup;
+        }
+
+
+        public void add(TextPositionSequence number) {
+
+            if (numberClusters.getElements().contains(number)) {
+                return;
+            }
+
+            numberClusters.addElement(number);
+            for (TextPositionSequence element : numberClusters.getElements()) {
+                if (matches(number, element, lookup)) {
+                    numberClusters.union(element, number);
+                }
+            }
+        }
+
+
+        public List<TextPositionSequence> getCurrentRightmostCluster() {
+
+            return numberClusters.getGroups()
+                    .stream()
+                    .filter(cluster -> cluster.size() > MINIMUM_MATCHES)
+                    .map(cluster -> cluster.stream()
+                            .sorted(new TextPositionSequenceComparator(lookup))
+                            .toList())
+                    .map(this::removeOutliers)
+//                    .map(this::filterByMinimumDensity)
+                    .filter(cluster -> cluster.size() > MINIMUM_MATCHES)
+                    .max(Comparator.comparingDouble(cluster -> cluster.get(0).getBBox().getMaxX())).orElse(Collections.emptyList());
+        }
+
+// does not seem to be doing much, ideally instead of using the height of the blocks i would like to use the height, beginning from the MainBody top,
+// but as the MainBody is often wrong, this results in some numbers being discarded even though they are correct.
+//        private List<TextPositionSequence> filterByMinimumDensity(List<TextPositionSequence> numbers) {
+//
+//            Map<ClassificationPage, List<TextPositionSequence>> clustersPerPage = numbers.stream()
+//                    .collect(Collectors.groupingBy(number -> lookup.get(number).page()));
+//
+//            List<TextPositionSequence> result = new ArrayList<>(numbers.size());
+//            clustersPerPage.keySet()
+//                    .stream()
+//                    .sorted(Comparator.comparingInt(ClassificationPage::getPageNumber))
+//                    .forEach(page -> {
+//                        var numbersOnPage = clustersPerPage.get(page);
+//
+//                        double height = numbersOnPage.stream()
+//                                .map(BoundingBox::getBBox)
+//                                .collect(RectangleTransformations.collectBBox()).getHeight();
+//
+//                        double count = numbersOnPage.size();
+//
+//                        if ((count / height) >= (DENSITY_THRESHOLD_COUNT / page.getPageHeight())) {
+//                            result.addAll(numbers);
+//                        }
+//                    });
+//            return result;
+//        }
+
+
+        public List<TextPositionSequence> removeOutliers(List<TextPositionSequence> numbers) {
+
+            List<TextPositionSequence> result = new ArrayList<>();
+
+            result.add(numbers.get(0));
+
+            for (int i = 1; i < numbers.size() - 1; i++) {
+                int prev = getNumberAsInt(numbers, i - 1);
+                int curr = getNumberAsInt(numbers, i);
+                int next = getNumberAsInt(numbers, i + 1);
+
+                if (!(curr <= prev || curr >= next) || !isBetterWithout(numbers, i)) {
+                    result.add(numbers.get(i));
+                }
+            }
+            if (getNumberAsInt(numbers, numbers.size() - 1) >= getNumberAsInt(numbers, Math.max(0, numbers.size() - 2))) {
+                result.add(numbers.get(numbers.size() - 1));
+            }
+
+            return result;
+        }
+
+
+        // Helper method to check if removing the current number results in a better order
+        public static boolean isBetterWithout(List<TextPositionSequence> numbers, int i) {
+
+            if (i == 0 || i == numbers.size() - 1) {
+                return false;
+            }
+
+            int prev = getNumberAsInt(numbers, i - 1);
+            int curr = getNumberAsInt(numbers, i);
+            int next = getNumberAsInt(numbers, i + 1);
+
+            return (prev <= next) && (Math.abs(prev - next) < Math.abs(prev - curr) + Math.abs(curr - next));
+        }
+
+    }
+
+
+    private static int getNumberAsInt(List<TextPositionSequence> numbers, int i) {
+
+        return Integer.parseInt(numbers.get(i).toString());
+    }
+
+}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/mapper/DocumentDataMapper.java

@@ -1,6 +1,10 @@
 package com.knecon.fforesight.service.layoutparser.processor.services.mapper;
 
+import static com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentTextDataProto.AllDocumentTextData;
+import static com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentTextDataProto.DocumentTextData;
+
 import java.awt.geom.Rectangle2D;
+import java.util.ArrayList;
 import java.util.HashMap;
 import java.util.HashSet;
 import java.util.List;
@@ -9,11 +13,15 @@ import java.util.Set;
 import java.util.stream.Collectors;
 
 import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentPage;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentPositionData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentStructure;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentTextData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.LayoutEngine;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentPageProto.AllDocumentPages;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentPageProto.DocumentPage;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentPositionDataProto.AllDocumentPositionData;
+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.DocumentStructureProto.DocumentStructure;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentStructureWrapper;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.EntryDataProto.EntryData;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.LayoutEngineProto.LayoutEngine;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.DocumentTree;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Document;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.DuplicatedParagraph;
@@ -32,44 +40,57 @@ public class DocumentDataMapper {
     public DocumentData toDocumentData(Document document) {
 
         List<DocumentTextData> documentTextData = document.streamTerminalTextBlocksInOrder()
-                .flatMap(textBlock -> textBlock.getAtomicTextBlocks().stream())
+                .flatMap(textBlock -> textBlock.getAtomicTextBlocks()
+                        .stream())
                 .distinct()
                 .map(DocumentDataMapper::toAtomicTextBlockData)
                 .toList();
 
+        AllDocumentTextData allDocumentTextData = AllDocumentTextData.newBuilder().addAllDocumentTextData(documentTextData).build();
+
         List<DocumentPositionData> atomicPositionBlockData = document.streamTerminalTextBlocksInOrder()
-                .flatMap(textBlock -> textBlock.getAtomicTextBlocks().stream())
+                .flatMap(textBlock -> textBlock.getAtomicTextBlocks()
+                        .stream())
                 .distinct()
                 .map(DocumentDataMapper::toAtomicPositionBlockData)
                 .toList();
 
-        Set<Long> nonEmptyTextBlocks = documentTextData.stream().mapToLong(DocumentTextData::getId).boxed().collect(Collectors.toSet());
+        AllDocumentPositionData allDocumentPositionData = AllDocumentPositionData.newBuilder().addAllDocumentPositionData(atomicPositionBlockData).build();
 
-        List<DocumentPage> documentPageData = document.getPages().stream().map(DocumentDataMapper::toPageData).toList();
-        DocumentStructure tableOfContentsData = toDocumentTreeData(document.getDocumentTree());
+        Set<Long> nonEmptyTextBlocks = documentTextData.stream()
+                .mapToLong(DocumentTextData::getId).boxed()
+                .collect(Collectors.toSet());
+
+        List<DocumentPage> documentPageData = document.getPages()
+                .stream()
+                .map(DocumentDataMapper::toPageData)
+                .toList();
+
+        AllDocumentPages allDocumentPages = AllDocumentPages.newBuilder().addAllDocumentPages(documentPageData).build();
+        DocumentStructureWrapper tableOfContentsData = toDocumentTreeData(document.getDocumentTree());
         return DocumentData.builder()
-                .documentTextData(documentTextData.toArray(new DocumentTextData[0]))
-                .documentPositions(atomicPositionBlockData.toArray(new DocumentPositionData[0]))
-                .documentPages(documentPageData.toArray(new DocumentPage[0]))
-                .documentStructure(tableOfContentsData)
+                .documentTextData(allDocumentTextData)
+                .documentPositions(allDocumentPositionData)
+                .documentPages(allDocumentPages)
+                .documentStructureWrapper(tableOfContentsData)
                 .build();
     }
 
 
-    private DocumentStructure toDocumentTreeData(DocumentTree documentTree) {
+    private DocumentStructureWrapper toDocumentTreeData(DocumentTree documentTree) {
 
-        return new DocumentStructure(toEntryData(documentTree.getRoot()));
+        return new DocumentStructureWrapper(DocumentStructure.newBuilder().setRoot(toEntryData(documentTree.getRoot())).build());
     }
 
 
-    private DocumentStructure.EntryData toEntryData(DocumentTree.Entry entry) {
+    private EntryData toEntryData(DocumentTree.Entry entry) {
 
-        Long[] atomicTextBlocks;
+        List<Long> atomicTextBlocks;
 
         if (entry.getNode().isLeaf()) {
             atomicTextBlocks = toAtomicTextBlockIds(entry.getNode().getLeafTextBlock());
         } else {
-            atomicTextBlocks = new Long[]{};
+            atomicTextBlocks = new ArrayList<>();
         }
 
         Map<String, String> properties = switch (entry.getType()) {
@@ -81,77 +102,81 @@ public class DocumentDataMapper {
             default -> new HashMap<>();
         };
 
-        DocumentStructure.EntryData.EntryDataBuilder documentBuilder = DocumentStructure.EntryData.builder()
-                .treeId(toPrimitiveIntArray(entry.getTreeId()))
-                .children(entry.getChildren().stream().map(DocumentDataMapper::toEntryData).toList())
-                .type(entry.getType())
-                .atomicBlockIds(atomicTextBlocks)
-                .pageNumbers(entry.getNode().getPages().stream().map(Page::getNumber).map(Integer::longValue).toArray(Long[]::new))
-                .properties(properties);
+        var documentBuilder = EntryData.newBuilder()
+                .addAllTreeId(entry.getTreeId())
+                .addAllChildren(entry.getChildren()
+                                        .stream()
+                                        .map(DocumentDataMapper::toEntryData)
+                                        .toList())
+                .setType(entry.getType())
+                .addAllAtomicBlockIds(atomicTextBlocks)
+                .addAllPageNumbers(entry.getNode().getPages()
+                                           .stream()
+                                           .map(Page::getNumber)
+                                           .map(Integer::longValue)
+                                           .toList())
+                .putAllProperties(properties);
         if (entry.getNode() != null) {
-            documentBuilder.engines(entry.getNode().getEngines());
+            documentBuilder.addAllEngines(entry.getNode().getEngines());
         } else {
-            documentBuilder.engines(new HashSet<>(Set.of(LayoutEngine.ALGORITHM)));
+            documentBuilder.addAllEngines(new HashSet<>(Set.of(LayoutEngine.ALGORITHM)));
         }
         return documentBuilder.build();
     }
 
 
-    private Long[] toAtomicTextBlockIds(TextBlock textBlock) {
+    private List<Long> toAtomicTextBlockIds(TextBlock textBlock) {
 
-        return textBlock.getAtomicTextBlocks().stream().map(AtomicTextBlock::getId).toArray(Long[]::new);
+        return textBlock.getAtomicTextBlocks()
+                .stream()
+                .map(AtomicTextBlock::getId)
+                .toList();
     }
 
 
     private DocumentPage toPageData(Page p) {
 
-        return DocumentPage.builder().rotation(p.getRotation()).height(p.getHeight()).width(p.getWidth()).number(p.getNumber()).build();
+        return DocumentPage.newBuilder().setRotation(p.getRotation()).setHeight(p.getHeight()).setWidth(p.getWidth()).setNumber(p.getNumber()).build();
     }
 
 
     private DocumentTextData toAtomicTextBlockData(AtomicTextBlock atomicTextBlock) {
 
-        return DocumentTextData.builder()
-                .id(atomicTextBlock.getId())
-                .page(atomicTextBlock.getPage().getNumber().longValue())
-                .searchText(atomicTextBlock.getSearchText())
-                .numberOnPage(atomicTextBlock.getNumberOnPage())
-                .start(atomicTextBlock.getTextRange().start())
-                .end(atomicTextBlock.getTextRange().end())
-                .lineBreaks(toPrimitiveIntArray(atomicTextBlock.getLineBreaks()))
+        return DocumentTextData.newBuilder()
+                .setId(atomicTextBlock.getId())
+                .setPage(atomicTextBlock.getPage().getNumber().longValue())
+                .setSearchText(atomicTextBlock.getSearchText())
+                .setNumberOnPage(atomicTextBlock.getNumberOnPage())
+                .setStart(atomicTextBlock.getTextRange().start())
+                .setEnd(atomicTextBlock.getTextRange().end())
+                .addAllLineBreaks(atomicTextBlock.getLineBreaks())
                 .build();
     }
 
 
     private DocumentPositionData toAtomicPositionBlockData(AtomicTextBlock atomicTextBlock) {
 
-        return DocumentPositionData.builder()
-                .id(atomicTextBlock.getId())
-                .positions(toPrimitiveFloatMatrix(atomicTextBlock.getPositions()))
-                .stringIdxToPositionIdx(toPrimitiveIntArray(atomicTextBlock.getStringIdxToPositionIdx()))
+        return DocumentPositionData.newBuilder()
+                .setId(atomicTextBlock.getId())
+                .addAllPositions(toPositions(atomicTextBlock.getPositions()))
+                .addAllStringIdxToPositionIdx(atomicTextBlock.getStringIdxToPositionIdx())
                 .build();
     }
 
 
-    private static float[][] toPrimitiveFloatMatrix(List<Rectangle2D> positions) {
+    private static List<Position> toPositions(List<Rectangle2D> rects) {
 
-        float[][] positionMatrix = new float[positions.size()][];
-        for (int i = 0; i < positions.size(); i++) {
-            positionMatrix[i] = toArray(positions.get(i));
+        List<Position> positions = new ArrayList<>();
+        for (Rectangle2D rect : rects) {
+            positions.add(toPosition(rect));
         }
-        return positionMatrix;
+        return positions;
     }
 
 
-    private static float[] toArray(Rectangle2D positions) {
+    private static Position toPosition(Rectangle2D rect) {
 
-        return new float[]{(float) positions.getMinX(), (float) positions.getMinY(), (float) positions.getWidth(), (float) positions.getHeight()};
-    }
-
-
-    private int[] toPrimitiveIntArray(List<Integer> list) {
-
-        return list.stream().mapToInt(Integer::intValue).toArray();
+        return Position.newBuilder().addValue((float) rect.getMinX()).addValue((float) rect.getMinY()).addValue((float) rect.getWidth()).addValue((float) rect.getHeight()).build();
     }
 
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/mapper/DocumentGraphMapper.java

@@ -8,10 +8,10 @@ import java.util.Map;
 import java.util.NoSuchElementException;
 
 import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentPage;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentPositionData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentStructure;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentTextData;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentPageProto.DocumentPage;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentPositionDataProto.AllDocumentPositionData;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentTextDataProto.AllDocumentTextData;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.EntryDataProto.EntryData;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.DocumentTree;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Document;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.DuplicatedParagraph;
@@ -41,27 +41,29 @@ public class DocumentGraphMapper {
         DocumentTree documentTree = new DocumentTree(document);
         Context context = new Context(documentData, documentTree);
 
-        context.pages.addAll(Arrays.stream(documentData.getDocumentPages())
+        context.pages.addAll(documentData.getDocumentPages().getDocumentPagesList()
+                                     .stream()
                                      .map(DocumentGraphMapper::buildPage)
                                      .toList());
 
-        context.documentTree.getRoot().getChildren().addAll(buildEntries(documentData.getDocumentStructure().getRoot().getChildren(), context));
+        context.documentTree.getRoot().getChildren().addAll(buildEntries(documentData.getDocumentStructure().getRoot().getChildrenList(), context));
 
         document.setDocumentTree(context.documentTree);
         document.setPages(new HashSet<>(context.pages));
-        document.setNumberOfPages(documentData.getDocumentPages().length);
+        document.setNumberOfPages(documentData.getDocumentPages().getDocumentPagesCount());
 
         document.setTextBlock(document.getTextBlock());
         return document;
     }
 
 
-    private List<DocumentTree.Entry> buildEntries(List<DocumentStructure.EntryData> entries, Context context) {
+    private List<DocumentTree.Entry> buildEntries(List<EntryData> entries, Context context) {
 
         List<DocumentTree.Entry> newEntries = new LinkedList<>();
-        for (DocumentStructure.EntryData entryData : entries) {
+        for (EntryData entryData : entries) {
 
-            List<Page> pages = Arrays.stream(entryData.getPageNumbers())
+            List<Page> pages = entryData.getPageNumbersList()
+                    .stream()
                     .map(pageNumber -> getPage(pageNumber, context))
                     .toList();
 
@@ -74,12 +76,12 @@ public class DocumentGraphMapper {
                 case FOOTER -> buildFooter(context);
                 case TABLE -> buildTable(context, entryData.getProperties());
                 case TABLE_CELL -> buildTableCell(context, entryData.getProperties());
-                case IMAGE -> buildImage(context, entryData.getProperties(), entryData.getPageNumbers());
+                case IMAGE -> buildImage(context, entryData.getProperties(), entryData.getPageNumbersList());
                 default -> throw new UnsupportedOperationException("Not yet implemented for type " + entryData.getType());
             };
 
-            if (entryData.getAtomicBlockIds().length > 0) {
-                TextBlock textBlock = toTextBlock(entryData.getAtomicBlockIds(), context, node);
+            if (entryData.getAtomicBlockIdsCount() > 0) {
+                TextBlock textBlock = toTextBlock(entryData.getAtomicBlockIdsList(), context, node);
                 node.setLeafTextBlock(textBlock);
                 switch (entryData.getType()) {
                     case HEADER -> pages.forEach(page -> page.setHeader((Header) node));
@@ -89,11 +91,10 @@ public class DocumentGraphMapper {
                             .forEach(atb -> atb.getPage().getTextBlocksOnPage().add(atb));
                 }
             }
-            List<Integer> treeId = Arrays.stream(entryData.getTreeId()).boxed()
-                    .toList();
+            List<Integer> treeId = entryData.getTreeIdList();
             node.setTreeId(treeId);
 
-            newEntries.add(DocumentTree.Entry.builder().treeId(treeId).children(buildEntries(entryData.getChildren(), context)).node(node).build());
+            newEntries.add(DocumentTree.Entry.builder().treeId(treeId).children(buildEntries(entryData.getChildrenList(), context)).node(node).build());
         }
         return newEntries;
     }
@@ -105,10 +106,10 @@ public class DocumentGraphMapper {
     }
 
 
-    private Image buildImage(Context context, Map<String, String> properties, Long[] pageNumbers) {
+    private Image buildImage(Context context, Map<String, String> properties, List<Long> pageNumbers) {
 
-        assert pageNumbers.length == 1;
-        Page page = getPage(pageNumbers[0], context);
+        assert pageNumbers.size() == 1;
+        Page page = getPage(pageNumbers.get(0), context);
         var builder = Image.builder();
         PropertiesMapper.parseImageProperties(properties, builder);
         return builder.documentTree(context.documentTree).page(page).build();
@@ -161,7 +162,7 @@ public class DocumentGraphMapper {
 
             DuplicatedParagraph duplicatedParagraph = DuplicatedParagraph.builder().documentTree(context.documentTree).build();
 
-            Long[] unsortedTextblockIds = PropertiesMapper.getUnsortedTextblockIds(properties);
+            var unsortedTextblockIds = PropertiesMapper.getUnsortedTextblockIds(properties);
             duplicatedParagraph.setUnsortedLeafTextBlock(toTextBlock(unsortedTextblockIds, context, duplicatedParagraph));
             return duplicatedParagraph;
 
@@ -171,9 +172,9 @@ public class DocumentGraphMapper {
     }
 
 
-    private TextBlock toTextBlock(Long[] atomicTextBlockIds, Context context, SemanticNode parent) {
+    private TextBlock toTextBlock(List<Long> atomicTextBlockIds, Context context, SemanticNode parent) {
 
-        return Arrays.stream(atomicTextBlockIds)
+        return atomicTextBlockIds.stream()
                 .map(atomicTextBlockId -> getAtomicTextBlock(context, parent, atomicTextBlockId))
                 .collect(new TextBlockCollector());
     }
@@ -181,10 +182,13 @@ public class DocumentGraphMapper {
 
     private AtomicTextBlock getAtomicTextBlock(Context context, SemanticNode parent, Long atomicTextBlockId) {
 
-        return AtomicTextBlock.fromAtomicTextBlockData(context.documentTextDataBlockData.get(Math.toIntExact(atomicTextBlockId)),
-                                                       context.atomicPositionBlockData.get(Math.toIntExact(atomicTextBlockId)),
+        return AtomicTextBlock.fromAtomicTextBlockData(context.documentTextDataBlockData.getDocumentTextDataList()
+                                                               .get(Math.toIntExact(atomicTextBlockId)),
+                                                       context.atomicPositionBlockData.getDocumentPositionDataList()
+                                                               .get(Math.toIntExact(atomicTextBlockId)),
                                                        parent,
-                                                       getPage(context.documentTextDataBlockData.get(Math.toIntExact(atomicTextBlockId)).getPage(), context));
+                                                       getPage(context.documentTextDataBlockData.getDocumentTextDataList()
+                                                                       .get(Math.toIntExact(atomicTextBlockId)).getPage(), context));
     }
 
 
@@ -207,18 +211,16 @@ public class DocumentGraphMapper {
 
         private final DocumentTree documentTree;
         private final List<Page> pages;
-        private final List<DocumentTextData> documentTextDataBlockData;
-        private final List<DocumentPositionData> atomicPositionBlockData;
+        private final AllDocumentTextData documentTextDataBlockData;
+        private final AllDocumentPositionData atomicPositionBlockData;
 
 
         Context(DocumentData documentData, DocumentTree documentTree) {
 
             this.documentTree = documentTree;
             this.pages = new LinkedList<>();
-            this.documentTextDataBlockData = Arrays.stream(documentData.getDocumentTextData())
-                    .toList();
-            this.atomicPositionBlockData = Arrays.stream(documentData.getDocumentPositions())
-                    .toList();
+            this.documentTextDataBlockData = documentData.getDocumentTextData();
+            this.atomicPositionBlockData = documentData.getDocumentPositions();
 
         }
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/mapper/PropertiesMapper.java

@@ -3,10 +3,11 @@ package com.knecon.fforesight.service.layoutparser.processor.services.mapper;
 import java.awt.geom.Rectangle2D;
 import java.util.Arrays;
 import java.util.HashMap;
+import java.util.List;
 import java.util.Locale;
 import java.util.Map;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentStructure;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentStructureWrapper;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.DuplicatedParagraph;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Image;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.ImageType;
@@ -20,11 +21,11 @@ public class PropertiesMapper {
     public static Map<String, String> buildImageProperties(Image image) {
 
         Map<String, String> properties = new HashMap<>();
-        properties.put(DocumentStructure.ImageProperties.IMAGE_TYPE, image.getImageType().toString());
-        properties.put(DocumentStructure.ImageProperties.TRANSPARENT, String.valueOf(image.isTransparent()));
-        properties.put(DocumentStructure.ImageProperties.POSITION, toString(image.getPosition()));
-        properties.put(DocumentStructure.ImageProperties.ID, image.getId());
-        properties.put(DocumentStructure.ImageProperties.REPRESENTATION_HASH, image.getRepresentationHash());
+        properties.put(DocumentStructureWrapper.ImageProperties.IMAGE_TYPE, image.getImageType().toString());
+        properties.put(DocumentStructureWrapper.ImageProperties.TRANSPARENT, String.valueOf(image.isTransparent()));
+        properties.put(DocumentStructureWrapper.ImageProperties.POSITION, toString(image.getPosition()));
+        properties.put(DocumentStructureWrapper.ImageProperties.ID, image.getId());
+        properties.put(DocumentStructureWrapper.ImageProperties.REPRESENTATION_HASH, image.getRepresentationHash());
         return properties;
     }
 
@@ -32,15 +33,19 @@ public class PropertiesMapper {
     public static Map<String, String> buildTableCellProperties(TableCell tableCell) {
 
         Map<String, String> properties = new HashMap<>();
-        properties.put(DocumentStructure.TableCellProperties.ROW, String.valueOf(tableCell.getRow()));
-        properties.put(DocumentStructure.TableCellProperties.COL, String.valueOf(tableCell.getCol()));
-        properties.put(DocumentStructure.TableCellProperties.HEADER, String.valueOf(tableCell.isHeader()));
+        properties.put(DocumentStructureWrapper.TableCellProperties.ROW, String.valueOf(tableCell.getRow()));
+        properties.put(DocumentStructureWrapper.TableCellProperties.COL, String.valueOf(tableCell.getCol()));
+        properties.put(DocumentStructureWrapper.TableCellProperties.HEADER, String.valueOf(tableCell.isHeader()));
 
         if (tableCell.getPages().size() > 1 || tableCell.getBBox().keySet().size() > 1) {
             throw new IllegalArgumentException("TableCell can only occur on a single page!");
         }
-        String bBoxString = toString(tableCell.getBBox().get(tableCell.getPages().stream().findFirst().get()));
-        properties.put(DocumentStructure.TableCellProperties.B_BOX, bBoxString);
+        String bBoxString = toString(tableCell.getBBox()
+                                             .get(tableCell.getPages()
+                                                          .stream()
+                                                          .findFirst()
+                                                          .get()));
+        properties.put(DocumentStructureWrapper.TableCellProperties.B_BOX, bBoxString);
 
         return properties;
     }
@@ -49,59 +54,62 @@ public class PropertiesMapper {
     public static Map<String, String> buildTableProperties(Table table) {
 
         Map<String, String> properties = new HashMap<>();
-        properties.put(DocumentStructure.TableProperties.NUMBER_OF_ROWS, String.valueOf(table.getNumberOfRows()));
-        properties.put(DocumentStructure.TableProperties.NUMBER_OF_COLS, String.valueOf(table.getNumberOfCols()));
+        properties.put(DocumentStructureWrapper.TableProperties.NUMBER_OF_ROWS, String.valueOf(table.getNumberOfRows()));
+        properties.put(DocumentStructureWrapper.TableProperties.NUMBER_OF_COLS, String.valueOf(table.getNumberOfCols()));
         return properties;
     }
 
 
     public static void parseImageProperties(Map<String, String> properties, Image.ImageBuilder builder) {
 
-        builder.imageType(parseImageType(properties.get(DocumentStructure.ImageProperties.IMAGE_TYPE)));
-        builder.transparent(Boolean.parseBoolean(properties.get(DocumentStructure.ImageProperties.TRANSPARENT)));
-        builder.position(DocumentStructure.parseRectangle2D(properties.get(DocumentStructure.ImageProperties.POSITION)));
+        builder.imageType(parseImageType(properties.get(DocumentStructureWrapper.ImageProperties.IMAGE_TYPE)));
+        builder.transparent(Boolean.parseBoolean(properties.get(DocumentStructureWrapper.ImageProperties.TRANSPARENT)));
+        builder.position(DocumentStructureWrapper.parseRectangle2D(properties.get(DocumentStructureWrapper.ImageProperties.POSITION)));
     }
 
 
     public static void parseTableCellProperties(Map<String, String> properties, TableCell.TableCellBuilder builder) {
 
-        builder.row(Integer.parseInt(properties.get(DocumentStructure.TableCellProperties.ROW)));
-        builder.col(Integer.parseInt(properties.get(DocumentStructure.TableCellProperties.COL)));
-        builder.header(Boolean.parseBoolean(properties.get(DocumentStructure.TableCellProperties.HEADER)));
-        builder.bBox(DocumentStructure.parseRectangle2D(properties.get(DocumentStructure.TableCellProperties.B_BOX)));
+        builder.row(Integer.parseInt(properties.get(DocumentStructureWrapper.TableCellProperties.ROW)));
+        builder.col(Integer.parseInt(properties.get(DocumentStructureWrapper.TableCellProperties.COL)));
+        builder.header(Boolean.parseBoolean(properties.get(DocumentStructureWrapper.TableCellProperties.HEADER)));
+        builder.bBox(DocumentStructureWrapper.parseRectangle2D(properties.get(DocumentStructureWrapper.TableCellProperties.B_BOX)));
     }
 
 
     public static void parseTableProperties(Map<String, String> properties, Table.TableBuilder builder) {
 
-        builder.numberOfRows(Integer.parseInt(properties.get(DocumentStructure.TableProperties.NUMBER_OF_ROWS)));
-        builder.numberOfCols(Integer.parseInt(properties.get(DocumentStructure.TableProperties.NUMBER_OF_COLS)));
+        builder.numberOfRows(Integer.parseInt(properties.get(DocumentStructureWrapper.TableProperties.NUMBER_OF_ROWS)));
+        builder.numberOfCols(Integer.parseInt(properties.get(DocumentStructureWrapper.TableProperties.NUMBER_OF_COLS)));
     }
 
 
     public static Map<String, String> buildDuplicateParagraphProperties(DuplicatedParagraph duplicatedParagraph) {
 
         Map<String, String> properties = new HashMap<>();
-        properties.put(DocumentStructure.DuplicateParagraphProperties.UNSORTED_TEXTBLOCK_ID, Arrays.toString(toAtomicTextBlockIds(duplicatedParagraph.getUnsortedLeafTextBlock())));
+        properties.put(DocumentStructureWrapper.DuplicateParagraphProperties.UNSORTED_TEXTBLOCK_ID,
+                       Arrays.toString(toAtomicTextBlockIds(duplicatedParagraph.getUnsortedLeafTextBlock())));
         return properties;
     }
 
 
     public static boolean isDuplicateParagraph(Map<String, String> properties) {
 
-        return properties.containsKey(DocumentStructure.DuplicateParagraphProperties.UNSORTED_TEXTBLOCK_ID);
+        return properties.containsKey(DocumentStructureWrapper.DuplicateParagraphProperties.UNSORTED_TEXTBLOCK_ID);
     }
 
 
-    public static Long[] getUnsortedTextblockIds(Map<String, String> properties) {
+    public static List<Long> getUnsortedTextblockIds(Map<String, String> properties) {
 
-        return toLongArray(properties.get(DocumentStructure.DuplicateParagraphProperties.UNSORTED_TEXTBLOCK_ID));
+        return toLongList(properties.get(DocumentStructureWrapper.DuplicateParagraphProperties.UNSORTED_TEXTBLOCK_ID));
     }
 
 
-    public static Long[] toLongArray(String ids) {
+    public static List<Long> toLongList(String ids) {
 
-        return Arrays.stream(ids.substring(1, ids.length() - 1).trim().split(",")).map(Long::valueOf).toArray(Long[]::new);
+        return Arrays.stream(ids.substring(1, ids.length() - 1).trim().split(","))
+                .map(Long::valueOf)
+                .toList();
     }
 
 
@@ -120,20 +128,23 @@ public class PropertiesMapper {
     public static String toString(Rectangle2D rectangle2D) {
 
         return String.format(Locale.US,
-                "%f%s%f%s%f%s%f",
-                rectangle2D.getX(),
-                DocumentStructure.RECTANGLE_DELIMITER,
-                rectangle2D.getY(),
-                DocumentStructure.RECTANGLE_DELIMITER,
-                rectangle2D.getWidth(),
-                DocumentStructure.RECTANGLE_DELIMITER,
-                rectangle2D.getHeight());
+                             "%f%s%f%s%f%s%f",
+                             rectangle2D.getX(),
+                             DocumentStructureWrapper.RECTANGLE_DELIMITER,
+                             rectangle2D.getY(),
+                             DocumentStructureWrapper.RECTANGLE_DELIMITER,
+                             rectangle2D.getWidth(),
+                             DocumentStructureWrapper.RECTANGLE_DELIMITER,
+                             rectangle2D.getHeight());
     }
 
 
     private static Long[] toAtomicTextBlockIds(TextBlock textBlock) {
 
-        return textBlock.getAtomicTextBlocks().stream().map(AtomicTextBlock::getId).toArray(Long[]::new);
+        return textBlock.getAtomicTextBlocks()
+                .stream()
+                .map(AtomicTextBlock::getId)
+                .toArray(Long[]::new);
     }
 
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/mapper/TaasDocumentDataMapper.java

@@ -6,7 +6,7 @@ import java.util.Locale;
 import java.util.concurrent.atomic.AtomicInteger;
 import java.util.stream.IntStream;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeTypeProto.NodeType;
 import com.knecon.fforesight.service.layoutparser.internal.api.data.taas.ParagraphData;
 import com.knecon.fforesight.service.layoutparser.internal.api.data.taas.Range;
 import com.knecon.fforesight.service.layoutparser.internal.api.data.taas.ResearchDocumentData;

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

@@ -5,6 +5,7 @@ import java.awt.geom.Point2D;
 import java.io.IOException;
 import java.util.ArrayList;
 import java.util.List;
+import java.util.Set;
 
 import org.apache.pdfbox.contentstream.operator.Operator;
 import org.apache.pdfbox.contentstream.operator.OperatorName;
@@ -50,6 +51,7 @@ import lombok.extern.slf4j.Slf4j;
 @Slf4j
 public class PDFLinesTextStripper extends PDFTextStripper {
 
+    private final static Set<String> DOT_LIKE_CHARACTERS = Set.of(".", "·", "•", "․", "‧", "∙", "⋅", "・", "．", "･", "…", "⸱", "﹒", "ꞏ");
     private final List<TextPositionSequence> textPositionSequences = new ArrayList<>();
     private final List<Ruling> rulings = new ArrayList<>();
     private final List<Ruling> graphicsPath = new ArrayList<>();
@@ -201,7 +203,7 @@ public class PDFLinesTextStripper extends PDFTextStripper {
 
         try {
             if (stroke && !getGraphicsState().getStrokingColor().isPattern() && isBlack(getGraphicsState().getStrokingColor()) || //
-                    !stroke && !getGraphicsState().getNonStrokingColor().isPattern() && isBlack(getGraphicsState().getNonStrokingColor())) {
+                !stroke && !getGraphicsState().getNonStrokingColor().isPattern() && isBlack(getGraphicsState().getNonStrokingColor())) {
                 rulings.addAll(path);
             }
         } catch (UnsupportedOperationException e) {
@@ -279,9 +281,9 @@ public class PDFLinesTextStripper extends PDFTextStripper {
                 startIndex = i;
             }
 
-            if (i > 0 && (textPositions.get(i).getUnicode().equals(" ") || textPositions.get(i).getUnicode().equals("\u00A0") || textPositions.get(i)
-                    .getUnicode()
-                    .equals("\t")) && i <= textPositions.size() - 2) {
+            if (i > 0
+                && (textPositions.get(i).getUnicode().equals(" ") || textPositions.get(i).getUnicode().equals("\u00A0") || textPositions.get(i).getUnicode().equals("\t"))
+                && i <= textPositions.size() - 2) {
                 List<TextPosition> sublist = textPositions.subList(startIndex, i);
                 if (checkIfSequenceContainsOnlyWhitespaces(sublist)) {
 
@@ -296,20 +298,31 @@ public class PDFLinesTextStripper extends PDFTextStripper {
                 }
                 startIndex = i + 1;
             }
+            if (isDottedLineFollowedByWord(textPositions, i, startIndex)) {
+                List<TextPosition> sublist = textPositions.subList(startIndex, i);
+                textPositionSequences.add(new TextPositionSequence(sublist, pageNumber, i == textPositions.size() - 1 && isParagraphStart));
+                startIndex = i;
+            }
+            if (isWordFollowedByDottedLine(textPositions, i, startIndex)) {
+                List<TextPosition> sublist = textPositions.subList(startIndex, i - 2);
+                textPositionSequences.add(new TextPositionSequence(sublist, pageNumber, i == textPositions.size() - 1 && isParagraphStart));
+                startIndex = i - 2;
+            }
         }
 
         List<TextPosition> sublist = textPositions.subList(startIndex, textPositions.size());
-        if (!sublist.isEmpty() && (sublist.get(sublist.size() - 1).getUnicode().equals(" ") || sublist.get(sublist.size() - 1)
-                .getUnicode()
-                .equals("\u00A0") || sublist.get(sublist.size() - 1).getUnicode().equals("\t"))) {
+        if (!sublist.isEmpty() && (sublist.get(sublist.size() - 1).getUnicode().equals(" ")
+                                   || sublist.get(sublist.size() - 1).getUnicode().equals("\u00A0")
+                                   || sublist.get(sublist.size() - 1).getUnicode().equals("\t"))) {
             sublist = sublist.subList(0, sublist.size() - 1);
         }
 
         if (!(sublist.isEmpty() || sublist.size() == 1 && (sublist.get(0).getUnicode().equals(" ") || sublist.get(0).getUnicode().equals("\u00A0") || sublist.get(0)
                 .getUnicode()
                 .equals("\t")))) {
-            if (previous != null && sublist.get(0).getYDirAdj() == previous.getYDirAdj() && sublist.get(0)
-                    .getXDirAdj() - (previous.getXDirAdj() + previous.getWidthDirAdj()) < 0.01) {
+            if (previous != null
+                && sublist.get(0).getYDirAdj() == previous.getYDirAdj()
+                && sublist.get(0).getXDirAdj() - (previous.getXDirAdj() + previous.getWidthDirAdj()) < 0.01) {
                 for (TextPosition t : sublist) {
                     textPositionSequences.get(textPositionSequences.size() - 1).add(t);
                 }
@@ -317,10 +330,43 @@ public class PDFLinesTextStripper extends PDFTextStripper {
                 textPositionSequences.add(new TextPositionSequence(sublist, pageNumber, isParagraphStart));
             }
         }
+
         super.writeString(text);
     }
 
 
+    private boolean isWordFollowedByDottedLine(List<TextPosition> textPositions, int i, int startIndex) {
+
+        return i - startIndex >= 4 //
+               && isDot(textPositions, i) //
+               && isDot(textPositions, i - 1) //
+               && isDot(textPositions, i - 2) //
+               && alphanumeric(textPositions, i - 3);
+    }
+
+
+    private static boolean isDottedLineFollowedByWord(List<TextPosition> textPositions, int i, int startIndex) {
+
+        return i - startIndex >= 4 //
+               && alphanumeric(textPositions, i)  //
+               && isDot(textPositions, i - 1) //
+               && isDot(textPositions, i - 2) //
+               && isDot(textPositions, i - 3);
+    }
+
+
+    private static boolean isDot(List<TextPosition> textPositions, int i) {
+
+        return DOT_LIKE_CHARACTERS.contains(textPositions.get(i).getUnicode());
+    }
+
+
+    private static boolean alphanumeric(List<TextPosition> textPositions, int i) {
+
+        return Character.isAlphabetic(textPositions.get(i).getUnicode().charAt(0)) || Character.isDigit(textPositions.get(i).getUnicode().charAt(0));
+    }
+
+
     public boolean checkIfCurrentPositionIsToTheRightOfPreviousPosition(int i, List<TextPosition> textPositions) {
 
         return i > 0 && textPositions.get(i).getXDirAdj() < textPositions.get(i - 1).getXDirAdj();
@@ -337,8 +383,9 @@ public class PDFLinesTextStripper extends PDFTextStripper {
 
     public boolean checkIfGapSizeBetweenCharactersSmallerThanMaximum(RedTextPosition previous, List<TextPosition> sublist, float maximumGapSize) {
 
-        return previous != null && sublist.get(0).getYDirAdj() == previous.getYDirAdj() && sublist.get(0)
-                .getXDirAdj() - (previous.getXDirAdj() + previous.getWidthDirAdj()) < maximumGapSize;
+        return previous != null
+               && sublist.get(0).getYDirAdj() == previous.getYDirAdj()
+               && sublist.get(0).getXDirAdj() - (previous.getXDirAdj() + previous.getWidthDirAdj()) < maximumGapSize;
     }
 
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/visualization/LayoutGridService.java

@@ -36,6 +36,9 @@ public class LayoutGridService {
         LayoutGrid layoutGrid = createLayoutGrid(document);
         Outline outline = OutlineMapper.createOutline(document);
         layoutGrid.setVisibleByDefault(layerVisibilityDefaultValue);
+
+        document.getLayoutDebugLayer().addSentenceVisualization(document.getTextBlock());
+
         if (document.getLayoutDebugLayer().isActive()) {
             viewerDocumentService.addLayerGroups(originFile, destinationFile, List.of(layoutGrid, document.getLayoutDebugLayer()), outline);
         } else {

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

@@ -1,5 +1,8 @@
 package com.knecon.fforesight.service.layoutparser.processor.utils;
 
+import static com.knecon.fforesight.service.layoutparser.processor.model.PageBlockType.FOOTER;
+import static com.knecon.fforesight.service.layoutparser.processor.model.PageBlockType.HEADER;
+
 import java.util.ArrayList;
 import java.util.HashMap;
 import java.util.List;
@@ -9,6 +12,7 @@ import java.util.stream.Collectors;
 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;
+import com.knecon.fforesight.service.layoutparser.processor.model.PageBlockType;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
 
 import lombok.experimental.UtilityClass;
@@ -26,35 +30,60 @@ public class HeaderFooterDetection {
 
     public boolean isLikelyFooter(TextPageBlock textPageBlock, ClassificationDocument document, ClassificationPage classificationPage) {
 
-        int numberOfPages = document.getPages().size();
-        if (numberOfPages < 3) {
-            // If the document has 1 or 2 pages this may lead to more false positives than actual findings.
-            return false;
-        }
-
-        int window = Math.min(numberOfPages, 8);
-
-        List<ClassificationPage> nearestPages = findNearestPages(classificationPage, document.getPages(), window);
-        List<List<AbstractPageBlock>> footerCandidates = getFooterCandidates(nearestPages);
-
-        return detectHeadersOrFootersByPageAssociation(textPageBlock.getText(), footerCandidates, window, footerWeights);
+        return isLikelyHeaderFooter(textPageBlock, document, classificationPage, FOOTER);
     }
 
 
     public boolean isLikelyHeader(TextPageBlock textPageBlock, ClassificationDocument document, ClassificationPage classificationPage) {
 
+        return isLikelyHeaderFooter(textPageBlock, document, classificationPage, HEADER);
+    }
+
+
+    private boolean isLikelyHeaderFooter(TextPageBlock textPageBlock, ClassificationDocument document, ClassificationPage classificationPage, PageBlockType type) {
+
         int numberOfPages = document.getPages().size();
         if (numberOfPages < 3) {
             // If the document has 1 or 2 pages this may lead to more false positives than actual findings.
             return false;
         }
 
+        List<TextPageBlock> textPageBlocks = classificationPage.getTextBlocks()
+                .stream()
+                .filter(TextPageBlock.class::isInstance)
+                .map(TextPageBlock.class::cast)
+                .collect(Collectors.toList());
+
+        if (textPageBlocks.isEmpty()) {
+            return false;
+        }
+
+        List<TextPageBlock> selectedBlocks;
+        if (type == HEADER) {
+            selectedBlocks = textPageBlocks.subList(0, Math.min(3, textPageBlocks.size()));
+        } else { //FOOTER
+            selectedBlocks = textPageBlocks.subList(Math.max(0, textPageBlocks.size() - 3), textPageBlocks.size());
+        }
+
+        if (!selectedBlocks.contains(textPageBlock)) {
+            // The textPageBlock is not among the selected blocks on its page
+            return false;
+        }
+
         int window = Math.min(numberOfPages, 8);
-
         List<ClassificationPage> nearestPages = findNearestPages(classificationPage, document.getPages(), window);
-        List<List<AbstractPageBlock>> headerCandidates = getHeaderCandidates(nearestPages);
 
-        return detectHeadersOrFootersByPageAssociation(textPageBlock.getText(), headerCandidates, window, headerWeights);
+        List<List<AbstractPageBlock>> candidates;
+        double[] weights;
+        if (type == HEADER) {
+            candidates = getHeaderCandidates(nearestPages);
+            weights = headerWeights;
+        } else { //FOOTER
+            candidates = getFooterCandidates(nearestPages);
+            weights = footerWeights;
+        }
+
+        return detectHeadersOrFootersByPageAssociation(textPageBlock.getText(), candidates, window, weights);
     }
 
 

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

@@ -14,7 +14,7 @@ import org.apache.pdfbox.pdmodel.PDPageContentStream;
 import org.apache.pdfbox.pdmodel.font.PDType1Font;
 import org.apache.pdfbox.pdmodel.font.Standard14Fonts;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
+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.processor.model.graph.nodes.Document;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Page;
@@ -117,6 +117,7 @@ public class PdfVisualisationUtility {
             case TABLE -> Color.ORANGE;
             case TABLE_CELL -> Color.GRAY;
             case IMAGE -> Color.MAGENTA;
+            case UNRECOGNIZED -> Color.PINK;
         }).build();
     }
 

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

@@ -0,0 +1,42 @@
+package com.knecon.fforesight.service.layoutparser.processor.utils;
+
+import java.io.BufferedOutputStream;
+import java.io.File;
+import java.io.FileOutputStream;
+import java.io.IOException;
+import java.util.zip.GZIPOutputStream;
+
+import com.google.protobuf.Message;
+import com.google.protobuf.MessageOrBuilder;
+import com.google.protobuf.Struct;
+import com.google.protobuf.util.JsonFormat;
+
+import lombok.SneakyThrows;
+import lombok.experimental.UtilityClass;
+
+@UtilityClass
+public class ProtobufUtil {
+
+    public static String toJson(MessageOrBuilder messageOrBuilder) throws IOException {
+        return JsonFormat.printer().print(messageOrBuilder);
+    }
+
+    @SuppressWarnings("unchecked")
+    public static Message fromJson(String json) throws IOException {
+        Message.Builder structBuilder = Struct.newBuilder();
+        JsonFormat.parser().ignoringUnknownFields().merge(json, structBuilder);
+        return structBuilder.build();
+    }
+
+
+    @SneakyThrows
+    public <T extends Message> File serializeToTempFile(T any) {
+        var tempFile = File.createTempFile("storage-protobuf", ".data");
+
+        try (var fos = new GZIPOutputStream(new BufferedOutputStream(new FileOutputStream(tempFile)))) {
+            any.writeTo(fos);
+            return tempFile;
+        }
+    }
+
+}

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

@@ -10,6 +10,7 @@ public final class TextNormalizationUtilities {
     public static final Pattern hyphenLineBreaks = Pattern.compile("[-~‐‒⁻−﹣゠⁓‑\\u00AD][\\r\\n]+");
     public static final Pattern linebreaks = Pattern.compile("[\\r\\n]+");
     public static final Pattern doubleWhitespaces = Pattern.compile("\\s{2,}");
+    public static final Pattern WHITESPACE_REMOVAL = Pattern.compile("\\s+");
 
 
     public String cleanString(String value) {
@@ -36,4 +37,11 @@ public final class TextNormalizationUtilities {
 
         return linebreaks.matcher(value).replaceAll(" ");
     }
+
+
+    public String removeAllWhitespaces(String value) {
+
+        return WHITESPACE_REMOVAL.matcher(value).replaceAll("");
+    }
+
 }

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

@@ -2,6 +2,7 @@ package com.knecon.fforesight.service.layoutparser.processor.utils;
 
 import java.awt.geom.Rectangle2D;
 import java.awt.geom.RectangularShape;
+import java.util.ArrayList;
 import java.util.Collection;
 import java.util.Comparator;
 import java.util.HashSet;
@@ -54,18 +55,26 @@ public class TextPositionOperations {
     private List<TextPositionSequence> sortUsingLineDetection(Set<TextPositionSequence> sequences) {
 
         return sortLines(groupByLine(sequences));
-
     }
 
 
     public List<TextPositionSequence> sortLines(Collection<Set<TextPositionSequence>> lines) {
 
-        return lines.stream()
-                .map(TextPositionOperations::sortByXDirAdj)
-                .filter(line -> !line.isEmpty())
-                .sorted(Comparator.comparing(line -> line.get(0), COMPARATOR_DIR_ADJ))
-                .flatMap(Collection::stream)
-                .toList();
+        List<List<TextPositionSequence>> lineBlocks = new ArrayList<>();
+        for (Set<TextPositionSequence> line : lines) {
+            List<TextPositionSequence> sortedLine = sortByXDirAdj(line);
+            if (!sortedLine.isEmpty()) {
+                lineBlocks.add(sortedLine);
+            }
+        }
+        // need to use old sorting, since COMPARATOR_DIR_ADJ is not transitive
+        QuickSort.sort(lineBlocks, Comparator.comparing(line -> line.get(0), COMPARATOR_DIR_ADJ));
+
+        List<TextPositionSequence> list = new ArrayList<>();
+        for (List<TextPositionSequence> textPositionSequences : lineBlocks) {
+            list.addAll(textPositionSequences);
+        }
+        return list;
     }
 
 
@@ -91,7 +100,7 @@ public class TextPositionOperations {
         for (TextPositionSequence sequence : sequences) {
             for (TextPositionSequence sequence2 : sequences) {
 
-                if (sequence.equals(sequence2) || unionFind.inSameSet(sequence, sequence2)) {
+                if (sequence.equals(sequence2)) { // || unionFind.inSameSet(sequence, sequence2)) doing this is actually slower than not doing it
                     continue;
                 }
 

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

@@ -5,9 +5,11 @@ import java.awt.geom.AffineTransform;
 import java.awt.geom.Line2D;
 import java.awt.geom.Point2D;
 import java.awt.geom.Rectangle2D;
+import java.text.BreakIterator;
 import java.util.Collection;
 import java.util.HashMap;
 import java.util.List;
+import java.util.Locale;
 import java.util.Map;
 import java.util.Set;
 import java.util.concurrent.atomic.AtomicInteger;
@@ -19,9 +21,12 @@ import com.knecon.fforesight.service.layoutparser.processor.docstrum.model.Bound
 import com.knecon.fforesight.service.layoutparser.processor.docstrum.model.Line;
 import com.knecon.fforesight.service.layoutparser.processor.docstrum.model.Zone;
 import com.knecon.fforesight.service.layoutparser.processor.model.AbstractPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.model.graph.TextRange;
+import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.TextBlock;
 import com.knecon.fforesight.service.layoutparser.processor.model.outline.OutlineObject;
 import com.knecon.fforesight.service.layoutparser.processor.model.table.CleanRulings;
 import com.knecon.fforesight.service.layoutparser.processor.model.table.Ruling;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.ListIdentifier;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPositionSequence;
 import com.knecon.fforesight.service.layoutparser.processor.utils.CoordinateTransforms;
 import com.knecon.fforesight.service.layoutparser.processor.utils.MarkedContentUtils;
@@ -76,7 +81,7 @@ public class LayoutDebugLayer extends LayoutDebugLayerConfig {
         visualizationsOnPage.getColoredLines()
                 .addAll(cleanRulings.buildAll()
                                 .stream()
-                                .map(ruling -> new ColoredLine(ruling, decideOnRulingColor(ruling), 0.5f))
+                                .map(ruling -> new ColoredLine(ruling, decideOnRulingColor(ruling), LINE_WIDTH))
                                 .toList());
     }
 
@@ -89,11 +94,34 @@ public class LayoutDebugLayer extends LayoutDebugLayerConfig {
         VisualizationsOnPage visualizationsOnPage = getOrCreateVisualizationsOnPage(pageNumber, this.rulings);
         visualizationsOnPage.getColoredLines()
                 .addAll(rulings.stream()
-                                .map(ruling -> new ColoredLine(ruling, decideOnRulingColor(ruling), 0.5f))
+                                .map(ruling -> new ColoredLine(ruling, decideOnRulingColor(ruling), LINE_WIDTH))
                                 .toList());
     }
 
 
+    public void addSentenceVisualization(TextBlock textBlock) {
+
+        if (!active) {
+            return;
+        }
+        AtomicInteger rotatingColorIdx = new AtomicInteger(0);
+        String text = textBlock.getSearchText();
+        BreakIterator sentenceIterator = BreakIterator.getSentenceInstance(Locale.ENGLISH);
+        sentenceIterator.setText(text);
+        int lastIdx = 0;
+        while (sentenceIterator.next() != BreakIterator.DONE) {
+            TextRange sentenceRange = new TextRange(lastIdx + textBlock.getTextRange().start(), sentenceIterator.current() + textBlock.getTextRange().start());
+            lastIdx = sentenceIterator.current();
+            Color color = getRotatingColor(rotatingColorIdx);
+            textBlock.getPositionsPerPage(sentenceRange)
+                    .forEach((page, bboxes) -> getOrCreateVisualizationsOnPage(page.getNumber(), this.sentences).getColoredRectangles()
+                            .addAll(bboxes.stream()
+                                            .map(bbox -> new ColoredRectangle(bbox, color, 1))
+                                            .toList()));
+        }
+    }
+
+
     private Color decideOnRulingColor(Ruling ruling) {
 
         return switch (ruling.getClassification()) {
@@ -155,7 +183,7 @@ public class LayoutDebugLayer extends LayoutDebugLayerConfig {
         visualizationsOnPage.getColoredRectangles()
                 .addAll(lines.stream()
                                 .map(BoundingBox::getBBoxPdf)
-                                .map(line -> new ColoredRectangle(line, LINES_COLOR, 0.5f))
+                                .map(line -> new ColoredRectangle(line, LINES_COLOR, LINE_WIDTH))
                                 .toList());
     }
 
@@ -171,7 +199,7 @@ public class LayoutDebugLayer extends LayoutDebugLayerConfig {
                                 .map(line -> line.stream()
                                         .map(BoundingBox::getBBoxPdf)
                                         .collect(RectangleTransformations.collectBBox()))
-                                .map(line -> new ColoredRectangle(line, LINES_COLOR, 0.5f))
+                                .map(line -> new ColoredRectangle(line, LINES_COLOR, LINE_WIDTH))
                                 .toList());
 
     }
@@ -248,7 +276,7 @@ public class LayoutDebugLayer extends LayoutDebugLayerConfig {
                 .map(Line::getCharacters)
                 .flatMap(Collection::stream)
                 .forEach(character -> {
-                    Color color = ROTATING_CHARACTER_COLOR.get(index.getAndIncrement() % ROTATING_CHARACTER_COLOR.size());
+                    Color color = getRotatingColor(index);
                     Rectangle2D charBBox = character.getTextPosition().getBBoxPdf();
                     characterVisualizations.getColoredRectangles().add(new ColoredRectangle(charBBox, color, 1));
                     character.getNeighbors()
@@ -263,6 +291,31 @@ public class LayoutDebugLayer extends LayoutDebugLayerConfig {
     }
 
 
+    public void addTocPages(List<TextPositionSequence> numbers, int page) {
+
+        if (!active) {
+            return;
+        }
+
+        VisualizationsOnPage visualizationsOnPage = getOrCreateVisualizationsOnPage(page, this.tocPages);
+        visualizationsOnPage.getColoredRectangles()
+                .addAll(numbers.stream()
+                                .map(BoundingBox::getBBoxPdf)
+                                .map(line -> new ColoredRectangle(line, LINES_COLOR, LINE_WIDTH))
+                                .toList());
+        visualizationsOnPage.getColoredRectangles()
+                .add(new ColoredRectangle(numbers.stream()
+                                                  .map(BoundingBox::getBBoxPdf)
+                                                  .collect(RectangleTransformations.collectBBox()), LINES_COLOR, LINE_WIDTH));
+    }
+
+
+    private static Color getRotatingColor(AtomicInteger index) {
+
+        return ROTATING_CHARACTER_COLOR.get(index.getAndIncrement() % ROTATING_CHARACTER_COLOR.size());
+    }
+
+
     public void addOutlineObjects(List<OutlineObject> outlineObjects, PageInformation pageInformation) {
 
         if (!active) {
@@ -299,4 +352,13 @@ public class LayoutDebugLayer extends LayoutDebugLayerConfig {
         visualizationsOnPage.getPlacedTexts().add(PlacedText.textFacingUp(outlineObject.getTitle(), textPoint, 10, outlineObject.isFound() ? Color.BLACK : Color.RED, FONT));
     }
 
+
+    public void addListIdentifiers(List<ListIdentifier> listIdentifiers) {
+
+        for (ListIdentifier listIdentifier : listIdentifiers) {
+            getOrCreateVisualizationsOnPage(listIdentifier.getPage(), this.listIdentifiers).getColoredRectangles()
+                    .add(new ColoredRectangle(listIdentifier.getWord().getBBoxPdf(), WORDS_COLOR, LINE_WIDTH));
+        }
+    }
+
 }

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

@@ -16,7 +16,7 @@ import java.util.stream.Collectors;
 import java.util.stream.Stream;
 
 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.internal.api.data.redaction.NodeTypeProto;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.DuplicatedParagraph;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Headline;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Image;
@@ -88,10 +88,19 @@ public class LayoutGrid extends LayoutGridLayerConfig {
     public void addTreeId(SemanticNode semanticNode) {
 
         Page page = semanticNode.getFirstPage();
-        if (semanticNode.getBBox().get(page) == null) {
+        if (semanticNode.getBBox()
+                    .get(page) == null) {
             return;
         }
-        addPlacedText(page, semanticNode.getBBox().get(page), semanticNode.getBBox().get(page), buildTreeIdString(semanticNode), 1, treeIds, TREEID_COLOR);
+        addPlacedText(page,
+                      semanticNode.getBBox()
+                              .get(page),
+                      semanticNode.getBBox()
+                              .get(page),
+                      buildTreeIdString(semanticNode),
+                      1,
+                      treeIds,
+                      TREEID_COLOR);
     }
 
 
@@ -116,11 +125,12 @@ public class LayoutGrid extends LayoutGridLayerConfig {
 
         Map<Page, Rectangle2D> bBoxMap = section.getBBox();
 
-        List<SemanticNode> subSections = section.streamAllSubNodesOfType(NodeType.SECTION)
+        List<SemanticNode> subSections = section.streamAllSubNodesOfType(NodeTypeProto.NodeType.SECTION)
                 .toList();
         Integer maxChildDepth = subSections.stream()
                 .map(node -> node.getTreeId().size())
-                .max(Integer::compareTo).orElse(section.getTreeId().size());
+                .max(Integer::compareTo)
+                .orElse(section.getTreeId().size());
         int ownDepth = section.getTreeId().size();
 
         Page firstPage = section.getFirstPage();
@@ -205,7 +215,8 @@ public class LayoutGrid extends LayoutGridLayerConfig {
         PlacedText newText = PlacedText.textFacingUp(s, upperLeftCorner, FONT_SIZE, color, FONT);
         float threshold = 1.5f * FONT_SIZE;
         Optional<PlacedText> conflictingText = placedTexts.stream()
-                .filter(pt -> Math.abs(pt.lineStart().getY() - newText.lineStart().getY()) <= threshold || Math.abs(pt.lineStart().getX() - newText.lineStart().getX()) <= threshold)
+                .filter(pt -> Math.abs(pt.lineStart().getY() - newText.lineStart().getY()) <= threshold
+                              && Math.abs(pt.lineStart().getX() - newText.lineStart().getX()) <= threshold)
                 .findFirst();
 
         if (conflictingText.isPresent()) {
@@ -286,7 +297,8 @@ public class LayoutGrid extends LayoutGridLayerConfig {
 
         List<ColoredLine> coloredLines = getOrCreateVisualizationsOnPage(page.getNumber(), sections).getColoredLines();
         int lineWidthModifier = maxChildDepth - ownDepth;
-        Rectangle2D r = RectangleTransformations.pad(semanticNode.getBBox().get(page), LINE_WIDTH * (1 + lineWidthModifier), LINE_WIDTH * (1 + lineWidthModifier));
+        Rectangle2D r = RectangleTransformations.pad(semanticNode.getBBox()
+                                                             .get(page), LINE_WIDTH * (1 + lineWidthModifier), LINE_WIDTH * (1 + lineWidthModifier));
 
         SemanticNode highestParent = semanticNode.getHighestParent();
         Rectangle2D highestParentRect = rectangleMap.get(new RectangleIdentifier(highestParent.getTreeId(), page.getNumber()));
@@ -335,7 +347,8 @@ public class LayoutGrid extends LayoutGridLayerConfig {
             List<Double> ys = yStream.collect(Collectors.toList());
             ys.remove(0);
 
-            Rectangle2D tableBBox = table.getBBox().get(page);
+            Rectangle2D tableBBox = table.getBBox()
+                    .get(page);
             List<ColoredLine> coloredLines = getOrCreateVisualizationsOnPage(page.getNumber(), tables).getColoredLines();
 
             xs.forEach(x -> {

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

@@ -29,7 +29,7 @@ dependencies {
     implementation(project(":layoutparser-service-processor"))
     implementation(project(":layoutparser-service-internal-api"))
 
-    implementation("com.iqser.red.commons:storage-commons:2.45.0")
+    implementation("com.iqser.red.commons:storage-commons:2.50.0")
     implementation("com.knecon.fforesight:tenant-commons:0.30.0")
     implementation("com.knecon.fforesight:tracing-commons:0.5.0")
     implementation("com.knecon.fforesight:lifecycle-commons:0.6.0")
@@ -38,15 +38,18 @@ dependencies {
     implementation("com.amazonaws:aws-java-sdk-s3:1.12.536")
     implementation("org.springframework.cloud:spring-cloud-starter-openfeign:4.0.4")
     implementation("net.logstash.logback:logstash-logback-encoder:7.4")
+    implementation("ch.qos.logback:logback-classic")
+
     implementation("com.pdftron:PDFNet:10.11.0")
 
     // for integration testing only
     testImplementation(project(":viewer-doc-processor"))
     testImplementation(project(":layoutparser-service-internal-api"))
+    testImplementation("com.google.protobuf:protobuf-java-util:4.27.1")
 
     testImplementation("org.springframework.boot:spring-boot-starter-amqp:${springBootStarterVersion}")
     testImplementation("org.springframework.boot:spring-boot-starter-test:${springBootStarterVersion}")
-    testImplementation("com.iqser.red.service:persistence-service-shared-api-v1:2.144.0")
+    testImplementation("com.iqser.red.service:persistence-service-shared-api-v1:2.564.0-RED9010.0")
     testImplementation("com.iqser.red.commons:jackson-commons:1.3.0")
     testImplementation("com.fasterxml.jackson.module:jackson-module-afterburner:${jacksonVersion}")
     testImplementation("com.fasterxml.jackson.datatype:jackson-datatype-jsr310:${jacksonVersion}")

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/processor/docstrum/model/BoundingBoxTest.java

@@ -0,0 +1,70 @@
+package com.knecon.fforesight.service.layoutparser.processor.docstrum.model;
+
+import static org.junit.jupiter.api.Assertions.*;
+
+import java.awt.geom.Rectangle2D;
+
+import org.junit.jupiter.api.Test;
+
+class BoundingBoxTest {
+
+    @Test
+    void testHorizontalDistance_NoOverlap() {
+
+        ConcreteBoundingBox box1 = new ConcreteBoundingBox(0, 0, 10, 10);
+        ConcreteBoundingBox box2 = new ConcreteBoundingBox(20, 0, 10, 10);
+
+        assertEquals(10, box1.horizontalDistance(box2));
+    }
+
+
+    @Test
+    void testHorizontalDistance_Overlap() {
+
+        ConcreteBoundingBox box1 = new ConcreteBoundingBox(0, 0, 10, 10);
+        ConcreteBoundingBox box2 = new ConcreteBoundingBox(5, 0, 10, 10);
+
+        assertEquals(0, box1.horizontalDistance(box2));
+    }
+
+
+    @Test
+    void testVerticalDistance_NoOverlap() {
+
+        ConcreteBoundingBox box1 = new ConcreteBoundingBox(0, 0, 10, 10);
+        ConcreteBoundingBox box2 = new ConcreteBoundingBox(0, 20, 10, 10);
+
+        assertEquals(10, box1.verticalDistance(box2));
+    }
+
+
+    @Test
+    void testVerticalDistance_Overlap() {
+
+        ConcreteBoundingBox box1 = new ConcreteBoundingBox(0, 0, 10, 10);
+        ConcreteBoundingBox box2 = new ConcreteBoundingBox(0, 5, 10, 10);
+
+        assertEquals(0, box1.verticalDistance(box2));
+    }
+
+
+    @Test
+    void testVerticalDistance_PartialOverlap() {
+
+        ConcreteBoundingBox box1 = new ConcreteBoundingBox(0, 0, 10, 10);
+        ConcreteBoundingBox box2 = new ConcreteBoundingBox(0, 8, 10, 10);
+
+        assertEquals(0, box1.verticalDistance(box2));
+    }
+
+
+    @Test
+    void testHorizontalDistance_PartialOverlap() {
+
+        ConcreteBoundingBox box1 = new ConcreteBoundingBox(0, 0, 10, 10);
+        ConcreteBoundingBox box2 = new ConcreteBoundingBox(8, 0, 10, 10);
+
+        assertEquals(0, box1.horizontalDistance(box2));
+    }
+
+}

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/processor/docstrum/model/ConcreteBoundingBox.java

@@ -0,0 +1,12 @@
+package com.knecon.fforesight.service.layoutparser.processor.docstrum.model;
+
+import java.awt.geom.Rectangle2D;
+
+class ConcreteBoundingBox extends BoundingBox {
+
+    ConcreteBoundingBox(double x, double y, double width, double height) {
+
+        this.bBox = new Rectangle2D.Double(x, y, width, height);
+    }
+
+}

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/server/HeadlinesGoldStandardIntegrationTest.java

@@ -8,6 +8,7 @@ import java.util.Set;
 import java.util.stream.Collectors;
 
 import org.assertj.core.api.Assertions;
+import org.junit.jupiter.api.Disabled;
 import org.junit.jupiter.api.Test;
 import org.junit.jupiter.api.extension.ExtendWith;
 import org.springframework.amqp.rabbit.core.RabbitAdmin;
@@ -27,7 +28,6 @@ import org.springframework.test.context.junit.jupiter.SpringExtension;
 
 import com.fasterxml.jackson.databind.ObjectMapper;
 import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.ChangeType;
-import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.RedactionLog;
 import com.iqser.red.storage.commons.StorageAutoConfiguration;
 import com.iqser.red.storage.commons.service.StorageService;
 import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingType;
@@ -46,6 +46,8 @@ import lombok.EqualsAndHashCode;
 import lombok.SneakyThrows;
 import lombok.ToString;
 
+
+@Disabled
 @ExtendWith(SpringExtension.class)
 @SpringBootTest(classes = Application.class, webEnvironment = SpringBootTest.WebEnvironment.RANDOM_PORT)
 @Import(HeadlinesGoldStandardIntegrationTest.RedactionIntegrationTestConfiguration.class)
@@ -105,13 +107,13 @@ public class HeadlinesGoldStandardIntegrationTest {
         ClassPathResource pdfFileResource = new ClassPathResource(filePath);
 
         Set<Headline> goldStandardHeadlines = new HashSet<>();
-        var goldStandardLog = objectMapper.readValue(redactionLogResource.getInputStream(), RedactionLog.class);
-        goldStandardLog.getRedactionLogEntry()
-                .removeIf(r -> !r.isRedacted() || r.getChanges()
-                        .get(r.getChanges().size() - 1).getType().equals(ChangeType.REMOVED));
-        goldStandardLog.getRedactionLogEntry()
-                .forEach(e -> goldStandardHeadlines.add(new Headline(e.getPositions()
-                                                                             .get(0).getPage(), e.getValue())));
+//        var goldStandardLog = objectMapper.readValue(redactionLogResource.getInputStream(), RedactionLog.class);
+//        goldStandardLog.getRedactionLogEntry()
+//                .removeIf(r -> !r.isRedacted() || r.getChanges()
+//                        .get(r.getChanges().size() - 1).getType().equals(ChangeType.REMOVED));
+//        goldStandardLog.getRedactionLogEntry()
+//                .forEach(e -> goldStandardHeadlines.add(new Headline(e.getPositions()
+//                                                                             .get(0).getPage(), e.getValue())));
 
         Document documentGraph = DocumentGraphFactory.buildDocumentGraph(LayoutParsingType.DOCUMINE,
                                                                          layoutParsingPipeline.parseLayout(LayoutParsingType.DOCUMINE,

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/server/LayoutparserEnd2EndTest.java

@@ -37,7 +37,7 @@ public class LayoutparserEnd2EndTest extends AbstractTest {
     @Disabled
     public void testLayoutParserEndToEnd() {
 
-        String filePath = "/home/kschuettler/Dokumente/TestFiles/NER Dataset/Syngenta prod/77c680315c31d403d2e023be023b2087.PREVIEW.pdf";
+        String filePath = "/home/kschuettler/Dokumente/LayoutparsingEvaluation/VV-340050.pdf";
 
         runForFile(filePath);
     }

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/server/OutlineDetectionTest.java

@@ -79,7 +79,7 @@ public class OutlineDetectionTest extends AbstractTest {
         var documentFile = new ClassPathResource(fileName).getFile();
 
         long start = System.currentTimeMillis();
-        ClassificationDocument classificationDocument = parseLayout(fileName, LayoutParsingType.REDACT_MANAGER_WITHOUT_DUPLICATE_PARAGRAPH);
+        ClassificationDocument classificationDocument = parseLayout(fileName, LayoutParsingType.DOCUMINE_OLD);
         Document document = buildGraph(fileName, classificationDocument);
         layoutGridService.addLayoutGrid(documentFile, document, new File(tmpFileName), true);
         OutlineObjectTree outlineObjectTree = classificationDocument.getOutlineObjectTree();
@@ -102,7 +102,7 @@ public class OutlineDetectionTest extends AbstractTest {
 
         TableOfContents tableOfContents = classificationDocument.getTableOfContents();
 
-        assertEquals(tableOfContents.getMainSections().size(), 10);
+        assertEquals(tableOfContents.getMainSections().size(), 9);
         assertEquals(tableOfContents.getMainSections().subList(1, 9)
                              .stream()
                              .map(tableOfContentItem -> sanitizeString(tableOfContentItem.getHeadline().toString()))
@@ -135,7 +135,7 @@ public class OutlineDetectionTest extends AbstractTest {
 
         List<SemanticNode> childrenOfTypeSectionOrSuperSection = document.getChildrenOfTypeSectionOrSuperSection();
 
-        assertEquals(childrenOfTypeSectionOrSuperSection.size(), 10);
+        assertEquals(childrenOfTypeSectionOrSuperSection.size(), 9);
         assertEquals(childrenOfTypeSectionOrSuperSection.subList(1, 9)
                              .stream()
                              .map(section -> sanitizeString(section.getHeadline().getLeafTextBlock().toString()))

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/server/SimplifiedTextServiceTest.java

@@ -46,8 +46,8 @@ public class SimplifiedTextServiceTest
         Document document = buildGraph(file);
         SimplifiedText simplifiedText = simplifiedSectionTextService.toSimplifiedText(document);
         List<SimplifiedSectionText> sectionTexts = simplifiedText.getSectionTexts();
-        assertThat(sectionTexts.stream().filter(section -> section.getText().equals(footerExample)).collect(Collectors.toList()).size()).isGreaterThan(0);
-        assertThat(sectionTexts.stream().filter(section -> section.getText().equals(headerExample)).collect(Collectors.toList()).size()).isGreaterThan(0);
+        assertThat(sectionTexts.stream().filter(section -> section.getText().contains(footerExample)).toList().size()).isGreaterThan(0);
+        assertThat(sectionTexts.stream().filter(section -> section.getText().contains(headerExample)).toList().size()).isGreaterThan(0);
 
 
 

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/server/graph/DocumentGraphJsonWritingTest.java

@@ -70,7 +70,7 @@ public class DocumentGraphJsonWritingTest extends BuildDocumentTest {
         var stem = Path.of("/tmp/DocumentGraphJsonWritingTest");
         stem.toFile().mkdirs();
         var tmpFilePath = stem.resolve(filename.getFileName());
-        mapper.writeValue(new FileOutputStream(new File(tmpFilePath + "_structure" + ".json")), documentData.getDocumentStructure());
+        mapper.writeValue(new FileOutputStream(new File(tmpFilePath + "_structure" + ".json")), documentData.getDocumentStructureWrapper());
         mapper.writeValue(new FileOutputStream(new File(tmpFilePath + "_text" + ".json")), documentData.getDocumentTextData());
         mapper.writeValue(new FileOutputStream(new File(tmpFilePath + "_positions" + ".json")), documentData.getDocumentPositions());
         mapper.writeValue(new FileOutputStream(new File(tmpFilePath + "_pages" + ".json")), documentData.getDocumentPages());

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/server/graph/DocumentGraphMappingTest.java

@@ -6,11 +6,13 @@ import static org.junit.jupiter.api.Assertions.assertTrue;
 import org.junit.jupiter.api.Test;
 
 import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentPage;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentPositionData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentStructure;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentTextData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
+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.DocumentStructureWrapper;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentTextDataProto.AllDocumentTextData;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.EntryDataProto.EntryData;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeTypeProto.NodeType;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Document;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Table;
 import com.knecon.fforesight.service.layoutparser.processor.services.mapper.DocumentDataMapper;
@@ -32,21 +34,21 @@ public class DocumentGraphMappingTest extends BuildDocumentTest {
         Document document = buildGraph(filename);
         DocumentData documentData = DocumentDataMapper.toDocumentData(document);
 
-        storageService.storeJSONObject(TenantContext.getTenantId(), filename + "_PAGES" + ".json", documentData.getDocumentPages());
-        storageService.storeJSONObject(TenantContext.getTenantId(), filename + "_TEXT" + ".json", documentData.getDocumentTextData());
-        storageService.storeJSONObject(TenantContext.getTenantId(), filename + "_POSITIONS" + ".json", documentData.getDocumentPositions());
-        storageService.storeJSONObject(TenantContext.getTenantId(), filename + "_STRUCTURE" + ".json", documentData.getDocumentStructure());
+        storageService.storeProtoObject(TenantContext.getTenantId(), filename + "_PAGES" + ".proto", documentData.getDocumentPages());
+        storageService.storeProtoObject(TenantContext.getTenantId(), filename + "_TEXT" + ".proto", documentData.getDocumentTextData());
+        storageService.storeProtoObject(TenantContext.getTenantId(), filename + "_POSITIONS" + ".proto", documentData.getDocumentPositions());
+        storageService.storeProtoObject(TenantContext.getTenantId(), filename + "_STRUCTURE" + ".proto", documentData.getDocumentStructure());
 
-        DocumentPage[] pageData = storageService.readJSONObject(TenantContext.getTenantId(), filename + "_PAGES" + ".json", DocumentPage[].class);
-        DocumentTextData[] atomicTextBlockData = storageService.readJSONObject(TenantContext.getTenantId(), filename + "_TEXT" + ".json", DocumentTextData[].class);
-        DocumentPositionData[] atomicPositionBlockData = storageService.readJSONObject(TenantContext.getTenantId(),
-                filename + "_POSITIONS" + ".json",
-                DocumentPositionData[].class);
-        DocumentStructure documentTreeData = storageService.readJSONObject(TenantContext.getTenantId(), filename + "_STRUCTURE" + ".json", DocumentStructure.class);
+        AllDocumentPages pageData = storageService.readProtoObject(TenantContext.getTenantId(), filename + "_PAGES" + ".proto", AllDocumentPages.parser());
+        AllDocumentTextData atomicTextBlockData = storageService.readProtoObject(TenantContext.getTenantId(), filename + "_TEXT" + ".proto", AllDocumentTextData.parser());
+        AllDocumentPositionData atomicPositionBlockData = storageService.readProtoObject(TenantContext.getTenantId(),
+                                                                                         filename + "_POSITIONS" + ".proto",
+                                                                                         AllDocumentPositionData.parser());
+        DocumentStructure documentTreeData = storageService.readProtoObject(TenantContext.getTenantId(), filename + "_STRUCTURE" + ".proto", DocumentStructure.parser());
 
         DocumentData documentData2 = DocumentData.builder()
                 .documentPages(pageData)
-                .documentStructure(documentTreeData)
+                .documentStructureWrapper(new DocumentStructureWrapper(documentTreeData))
                 .documentTextData(atomicTextBlockData)
                 .documentPositions(atomicPositionBlockData)
                 .build();
@@ -73,10 +75,9 @@ public class DocumentGraphMappingTest extends BuildDocumentTest {
 
     private static boolean allTablesHavePositiveNumberOfRowsAndColumns(DocumentData documentData) {
 
-        return documentData.getDocumentStructure()
-                .streamAllEntries()
+        return documentData.getDocumentStructureWrapper().streamAllEntries()
                 .filter(entryData -> entryData.getType().equals(NodeType.TABLE))
-                .map(DocumentStructure.EntryData::getProperties)
+                .map(EntryData::getPropertiesMap)
                 .map(properties -> {
                     var builder = Table.builder();
                     PropertiesMapper.parseTableProperties(properties, builder);

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/server/graph/DocumentReadingOrderTest.java

@@ -51,7 +51,7 @@ import lombok.extern.slf4j.Slf4j;
 @Slf4j
 public class DocumentReadingOrderTest extends BuildDocumentTest {
 
-    private static final boolean DRAW_DIR_ADJ_COORDS = true;
+    private static final boolean DRAW_DIR_ADJ_COORDS = false;
     public static final List<LayoutParsingType> LAYOUT_PARSING_TYPES = List.of(LayoutParsingType.DOCUMINE,
                                                                                LayoutParsingType.DOCUMINE_OLD,
                                                                                LayoutParsingType.REDACT_MANAGER,
@@ -82,7 +82,7 @@ public class DocumentReadingOrderTest extends BuildDocumentTest {
     @Disabled
     public void drawDirAdjForFile() {
 
-        String pdfFile = "/home/kschuettler/Dokumente/Ticket Related/RED-9974/026dc94b019bc2348a4c54f0c6c4516f.ORIGIN.pdf";
+        String pdfFile = "/home/kschuettler/Dokumente/TestFiles/OCR/VV-331340/VV-331340_OCRED_first15.pdf";
 
         ClassificationDocument classificationDocument = parseLayout(pdfFile, LayoutParsingType.DOCUMINE_OLD);
 

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/server/graph/TextRangeTest.java

@@ -76,9 +76,10 @@ class TextRangeTest {
         assertEquals(List.of(new TextRange(10, 40), new TextRange(40, 100)), startTextRange.split(List.of(40)));
         assertEquals(1, startTextRange.split(Collections.emptyList()).size());
         assertEquals(1, startTextRange.split(List.of(startTextRange.start())).size());
+        assertEquals(1, startTextRange.split(List.of(100)).size());
         assertThrows(IndexOutOfBoundsException.class, () -> startTextRange.split(Collections.singletonList(0)));
-        assertThrows(IndexOutOfBoundsException.class, () -> startTextRange.split(Collections.singletonList(100)));
-        assertThrows(IndexOutOfBoundsException.class, () -> startTextRange.split(List.of(12, 40, 100)));
+        assertThrows(IndexOutOfBoundsException.class, () -> startTextRange.split(Collections.singletonList(101)));
+        assertThrows(IndexOutOfBoundsException.class, () -> startTextRange.split(List.of(12, 40, 101)));
     }
 
 }

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/server/services/RulingCleaningServiceTest.java

@@ -17,8 +17,9 @@ import org.junit.jupiter.api.Test;
 import org.springframework.core.io.ClassPathResource;
 
 import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentStructure;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentStructureWrapper;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.EntryDataProto.EntryData;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeTypeProto.NodeType;
 import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingType;
 import com.knecon.fforesight.service.layoutparser.processor.model.PageContents;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Document;
@@ -119,7 +120,7 @@ public class RulingCleaningServiceTest extends BuildDocumentTest {
                         Map.of("file",filename.toFile().toString())));
         DocumentData documentDataBefore = DocumentDataMapper.toDocumentData(documentGraphBefore);
         DocumentData documentDataAfter = DocumentDataMapper.toDocumentData(documentGraphAfter);
-        if (!compareStructures(documentDataBefore.getDocumentStructure(), documentDataAfter.getDocumentStructure())) {
+        if (!compareStructures(documentDataBefore.getDocumentStructureWrapper(), documentDataAfter.getDocumentStructureWrapper())) {
             String tmpFileNameBefore = "/tmp/before." + filename.getFileName().toString();
             try (PDDocument pdDocument = Loader.loadPDF(filename.toFile())) {
                 PdfDraw.drawDocumentGraph(pdDocument, documentGraphBefore);
@@ -136,11 +137,11 @@ public class RulingCleaningServiceTest extends BuildDocumentTest {
 
 
     @SneakyThrows
-    private boolean compareStructures(DocumentStructure structure1, DocumentStructure structure2) {
+    private boolean compareStructures(DocumentStructureWrapper structure1, DocumentStructureWrapper structure2) {
 
         List<Table> listStructure1 = structure1.streamAllEntries()
                 .filter(entryData -> entryData.getType().equals(NodeType.TABLE))
-                .map(DocumentStructure.EntryData::getProperties)
+                .map(EntryData::getPropertiesMap)
                 .map(properties -> {
                     var builder = Table.builder();
                     PropertiesMapper.parseTableProperties(properties, builder);
@@ -150,7 +151,7 @@ public class RulingCleaningServiceTest extends BuildDocumentTest {
 
         List<Table> listStructure2 = structure2.streamAllEntries()
                 .filter(entryData -> entryData.getType().equals(NodeType.TABLE))
-                .map(DocumentStructure.EntryData::getProperties)
+                .map(EntryData::getPropertiesMap)
                 .map(properties -> {
                     var builder = Table.builder();
                     PropertiesMapper.parseTableProperties(properties, builder);

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/server/utils/visualizations/PdfDraw.java

@@ -17,7 +17,7 @@ import org.apache.pdfbox.pdmodel.font.Standard14Fonts;
 import org.apache.pdfbox.util.Matrix;
 import org.springframework.core.io.ClassPathResource;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
+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.processor.model.graph.nodes.Document;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Page;
@@ -233,6 +233,7 @@ public class PdfDraw {
             case TABLE -> Color.ORANGE;
             case TABLE_CELL -> Color.GRAY;
             case IMAGE -> Color.MAGENTA;
+            case UNRECOGNIZED -> Color.PINK;
         }).build();
     }
 

layoutparser-service/viewer-doc-processor/src/main/java/com/knecon/fforesight/service/viewerdoc/LayerIdentifier.java

@@ -55,7 +55,11 @@ public record LayerIdentifier(String name, String markedContentName) {
     public static final LayerIdentifier NEIGHBOURS = new LayerIdentifier("Neighbours", "NEIGHBOURS");
     public static final LayerIdentifier CHARACTERS = new LayerIdentifier("Characters", "CHARACTERS");
     public static final LayerIdentifier OUTLINE_OBJECTS = new LayerIdentifier("Outline Positions", "OUTLINE_OBJECTS");
+    public static final LayerIdentifier SENTENCES = new LayerIdentifier("Sentences", "SENTENCES");
+    public static final LayerIdentifier TOC_PAGES = new LayerIdentifier("TOC pages", "TOC_PAGES");
+    public static final LayerIdentifier LIST_IDENTIFIERS = new LayerIdentifier("List identifiers", "LIST_IDENTIFIERS");
 
+    // Visual layout parser
     public static final LayerIdentifier KNECON_VISUAL_PARSING = new LayerIdentifier("Visual Layout Parser", "VISUAL_PARSING");
 
     //ocr

layoutparser-service/viewer-doc-processor/src/main/java/com/knecon/fforesight/service/viewerdoc/layers/LayoutDebugLayerConfig.java

@@ -18,6 +18,8 @@ public class LayoutDebugLayerConfig extends AbstractLayerGroup {
 
     protected static final Standard14EmbeddableFont FONT = Standard14EmbeddableFont.helvetica();
 
+    public static final float LINE_WIDTH = 0.5f;
+
     protected static final Color WORDS_COLOR = new Color(68, 84, 147);
     protected static final Color LINES_COLOR = new Color(152, 45, 179);
     protected static final Color ZONES_COLOR = new Color(131, 38, 38);
@@ -55,6 +57,9 @@ public class LayoutDebugLayerConfig extends AbstractLayerGroup {
     protected final Visualizations neighbours = Visualizations.builder().layer(LayerIdentifier.NEIGHBOURS).build();
     protected final Visualizations characters = Visualizations.builder().layer(LayerIdentifier.CHARACTERS).build();
     protected final Visualizations outlineObjects = Visualizations.builder().layer(LayerIdentifier.OUTLINE_OBJECTS).build();
+    protected final Visualizations sentences = Visualizations.builder().layer(LayerIdentifier.SENTENCES).build();
+    protected final Visualizations tocPages = Visualizations.builder().layer(LayerIdentifier.TOC_PAGES).build();
+    protected final Visualizations listIdentifiers = Visualizations.builder().layer(LayerIdentifier.LIST_IDENTIFIERS).build();
 
 
     public List<Visualizations> getVisualizations() {
@@ -63,14 +68,18 @@ public class LayoutDebugLayerConfig extends AbstractLayerGroup {
                        neighbours,//
                        words, //
                        lines, //
+                       sentences, //
                        zones, //
                        rulings, //
                        clean_rulings, //
                        cells, //
                        mainBody, //
                        markedContent, //
-                       outlineObjects //
+                       outlineObjects, //
+                       tocPages, //
+                       listIdentifiers //
         );
     }
 
+
 }

layoutparser-service/viewer-doc-processor/src/main/java/com/knecon/fforesight/service/viewerdoc/layers/OcrDebugLayerConfig.java

@@ -25,10 +25,11 @@ public class OcrDebugLayerConfig extends AbstractLayerGroup {
 
     protected final Visualizations debugText = Visualizations.builder().layer(LayerIdentifier.KNECON_OCR_TEXT_DEBUG).visibleByDefault(true).build();
     protected final Visualizations tableLines = Visualizations.builder().layer(LayerIdentifier.KNECON_OCR_LINE_DEBUG).visibleByDefault(true).build();
-    protected final Visualizations overlappedText = Visualizations.builder().layer(LayerIdentifier.KNECON_OCR_OVERLAPPED_TEXT).visibleByDefault(true).build();
+    protected final Visualizations overlappedText = Visualizations.builder().layer(LayerIdentifier.KNECON_OCR_OVERLAPPED_TEXT).visibleByDefault(false).build();
     protected final Visualizations debugBBox = Visualizations.builder().layer(LayerIdentifier.KNECON_OCR_BBOX_DEBUG).visibleByDefault(false).build();
 
 
+
     @Override
     public List<Visualizations> getVisualizations() {
 
@@ -36,10 +37,4 @@ public class OcrDebugLayerConfig extends AbstractLayerGroup {
     }
 
 
-    @Override
-    public boolean isVisibleByDefault() {
-
-        return true;
-    }
-
 }

layoutparser-service/viewer-doc-processor/src/main/java/com/knecon/fforesight/service/viewerdoc/service/OutlineUtility.java

@@ -68,10 +68,10 @@ public class OutlineUtility {
     public static void deleteExistingOutline(PDFDoc doc) {
 
         Bookmark firstBookmark = doc.getFirstBookmark();
-//        while (firstBookmark != null && firstBookmark.isValid()) {
+        while (firstBookmark != null && firstBookmark.isValid()) {
             firstBookmark.delete();
             firstBookmark = doc.getFirstBookmark();
-//        }
+        }
 
     }